trends in telecommunication reform 2006

Transcription

Te l e c o m m u n i c a t i o n
U n i o n
2 020060 6
I n t e r n a t i o n a l
N
I
S
D
TRECONMMUNICATION
*28388*
Printed in Switzerland
Geneva, 2006
ISBN 92-61-11431-8
T R E N D S I N T E L E C O M M U N I C AT I O N R E F O R M
TELE
m
r
o
f
re
6
0
20
G
N
I
T
A
L
REGU
E
H
T
I N ADBAND
BRO
D
L
R
WO
International
Union
ALSO AVAILABLE FROM ITU
PUBLICATIONS
Trends in Telecommunication Reform: Licensing in an Era of Convergence, 2004/05 (6th Edition) ..............................95 CHF
Trends in Telecommunication Reform: Promoting Universal Access to ICTs, 2003 (5th Edition) ..................................90 CHF
Trends in Telecommunication Reform: Effective Regulation, 2002 (4th Edition) ...........................................................90 CHF
Trends in Telecommunication Reform: Interconnection Regulation, 2000-2001 (3rd Edition) .......................................90 CHF
Trends in Telecommunication Reform: Convergence and Regulation, 1999 (2nd Edition) .............................................75 CHF
General Trends in Telecommunication Reform 1998: World Volume I...........................................................................75 CHF
General Trends in Telecommunication Reform 1998: Africa Volume II ..........................................................................65 CHF
General Trends in Telecommunication Reform 1998: Americas Volume III ...................................................................55 CHF
General Trends in Telecommunication Reform 1998: Arab States Volume IV ...............................................................45 CHF
General Trends in Telecommunication Reform 1998: Asia Pacific Volume V ................................................................60 CHF
General Trends in Telecommunication Reform 1998: Europe Volume VI .......................................................................72 CHF
Collection of five Regional reports (Volumes II-VI) ......................................................................................................297 CHF
Collection of Regional and World reports (Volumes I-VI).............................................................................................372 CHF
The Arab Book: Telecommunication Policies for the Arab Region, 2002 ......................................................................50 CHF
The Blue Book: Telecommunication Policies for the Americas, 2000 (2nd Edition) ......................................................50 CHF
Please contact the ITU Sales Service:
Tel.:
+41 22 730 5111
Fax:
+41 22 730 5194
E-mail:
[email protected]
Website: www.itu.int
© ITU 2006
International Telecommunication Union
Place des Nations
CH-1211 Geneva, Switzerland
First printing 2006
All rights reserved. No part of this publication may be reproduced, by any means whatsoever, without the prior written
permission of the International Telecommunication Union.
Denominations and classifications employed in this publication do not imply any opinion on the part of the International
Telecommunication Union concerning the legal or other status of any territory or any endorsement or acceptance of any boundary.
Where the designation “country” appears in this publication, it covers countries and territories.
ISBN 92-61-11431-8
Note: Discounts are available for all ITU Member States, Sector Members and Least Developed Countries.
U n i o n
N
I
S
D
TRECONMMUNICATION
TELE
m
r
o
f
re
02 06
G
N
I
T
A
L
REGU
E
H
T
I N ADBAND
BRO
v1.01
D
L
R
WO
International
Union
This report was prepared by a team led by Doreen Bogdan-Martin and Susan Schorr. The team included: John Alden,
William Bratton, Tracy Cohen, Yang-Soon Lee, Olli Mattila, Michael Best, John Muleta, John Palfrey, Bjorn Pehrson, Audrey Selian,
Wu Wei Shi, Russell Southwood and Nancy Sundberg. We specially thank Tania Bezago, our GREX Advisor, who provided additional
research and Esperanza Magpantay, BDT, who contributed ICT statistics.
The cover was designed by Stéphane Rollet.
The report was edited by John Alden. The regulatory tables were prepared by Kevin Munn and Nancy Sundberg. The ITU
Publications Composition Department was responsible for production of the report.
The report has benefited from the comments and advice of ITU staff including: Tim Kelly, Colin Langtry, Fabio Leite, Ricardo
Passerini and Robert Shaw.
We would especially like to thank for their comments: Martin Cave (Warwick Business School, University of Warwick);
Claudia Sarrocco (Organisation for Economic Co-operation and Development); Suresh Ramasubramanian (Outblaze); John Haydon
(Australian Communications and Media Authority); Tom Dale (Australian Department of Communications, IT and the Arts); Ewan
Sutherland (International Telecommunication Users Group); Derek Bambauer, David Abrams, Jonathan Zittrain (the Berkman Center);
and César David Moliné (Indotel, Dominican Republic).
These contributions, together with the support from ICT ministries and regulators, and others who have provided data and
background material, are gratefully acknowledged. Without their support, a report of this nature would be impossible.
ii
FOREWORD
We are proud to present the seventh edition of
Trends in Telecommunication Reform, an integral part of
our dialogue with the world’s information and communications technology (ICT) policy-makers and
regulators. This 7th edition has been released at a
time of remarkable transformation of the information and communication technology (ICT) sector,
fuelled by a combination of technological, market,
policy and regulatory developments. These changes
include unparalleled numbers of voice telephone
subscribers, the rise of IP-enabled networks and
Voice over IP (VoIP) services, initial-yet promisingdeployment of fixed line broadband and broadband
wireless access (BWA) services and intelligent radio
devices. At the same time that developed countries
are busy planning for the deployment of next generation networks and visualize a world of ubiquitous
networks, most developing countries have expanded
their continuing quest to provide universal access
to basic voice services to include universal access to
broadband internet services. Are developing countries making any progress in this quest? How can
regulators harness the potential of new technologies
and innovative business models to foster ICT sector
development?
In 1984, the Maitland Commission’s ‘Missing
Link’ report identified the challenge of bringing basic telecommunication services within easy reach of
all the world’s people by the early part of the 21st
century. The Missing Link report has been the
touchstone for the work of the Telecommunication
Development Bureau (BDT) of the International
Telecommunication Union (ITU) since its inception. In 2002, mobile cellular communications were
heralded as the answer to the missing link. Second
generation mobile services have been rolled out
to more people in the developing world than the
Maitland Commission could ever have imagined
possible-although the challenge of universal access
to basic services remains. The success of mobile
communications is linked to policy and regulatory
reforms as well as innovative business models and
new technological applications, such as short message services (SMS).
Twenty years later, the World Summit on the Information Society (WSIS) set even more ambitious
targets, to extend the internet to all the world’s villages by 2015 as the foundation for building the In-
formation Society. The WSIS also identified the vital
role the regulatory framework plays in enabling the
Information Society. The global community of national communications regulators met on the eve of
the second phase of WSIS to develop a new vision
of a regulatory framework to promote the deployment of broadband internet communication services
worldwide, in developing and developed countries
alike. At that same meeting, known as the ITU Global Symposium for Regulators (GSR), regulators
recognized that full participation in the Information
Society requires access to broadband internet services.
Low cost technologies exist today that can promote broadband access and enable developing countries to “leapfrog” over older technologies to advance
into the broadband future rapidly. Many businesses
stand ready to start providing broadband services in
developing countries-mirroring the keen interest
second-generation mobile service providers demonstrated in developing countries throughout this
decade. Although broadband technologies and business models hold great promise, the pace of broadband take-up hinges on the regulatory framework.
In many countries, today’s broadband ‘missing link’
is the regulatory framework. Regulators have an
unprecedented opportunity to speed the uptake of
broadband to enable the Information Society. Today’s broadband challenge requires new thinking,
and an end to business as usual. This publication is
designed to enable regulators and policy-makers to
meet this challenge.
The report has been prepared by the ITU’s Telecommunication Development Bureau. The authors
have benefited from comments and input from a
range of people, inside and outside of ITU. The
views expressed in the report, however, are those of
the authors and do not necessarily reflect the opinions of ITU or its members.
Hamadoun I. Touré
Director
Telecommunication Development Bureau
7 March 2006
iii
CONTENTS
1
1.9
MARKET AND REGULATORY TRENDS IN THE ICT
SECTOR ...............................................................................1
What will it take to bring broadband to the masses? ............... 1
ICT infrastructure deployment ................................................ 2
Market Trends in Privatizations, Investments and Services....... 6
Competition .......................................................................... 9
National Regulatory Authorities, Who Rules? ........................ 12
Regulatory Challenges of VoIP ............................................ 14
Spectrum Management ....................................................... 16
Dealing with Spam ............................................................. 16
Conclusion ......................................................................... 18
2
2.1
2.2
2.3
WHAT IS “BROADBAND”? ............................................ 21
What Do We Mean When We Say “Broadband”? ................... 21
A Short Broadband Taxonomy ............................................. 22
Conclusion ......................................................................... 26
3
3.1
3.2
3.3
3.4
3.5
3.6
UNDERSTANDING BROADBAND TECHNOLOGIES ..... 27
How Broadband Networks Are Designed ............................... 27
A Wire-line Broadband Roadmap ......................................... 29
Broadband Wireless Access (BWA) Networks ....................... 33
A Decision Framework for Broadband................................... 41
Power requirements for broadband ....................................... 44
Conclusion ......................................................................... 49
4
THE ROLE OF THE REGULATOR IN BROADBAND
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
iv
P
A
G
E
P
A
G
E
DEVELOPMENT ............................................................... 51
Introduction ........................................................................ 51
The Importance of Broadband in Developing Countries ......... 54
Key Issues in Promoting Broadband in Developing Countries 54
Providing Incentives for Network Investment......................... 56
Broadband Licensing .......................................................... 58
Alternative Approaches to Broadband Deployment ............... 60
Competition and Industry Regulation.................................... 63
Increasing Broadband Awareness ......................................... 69
Conclusion ......................................................................... 71
5
5.1
5.2
5.3
5.4
5.5
5.6
5.7
BROADBAND SPECTRUM MANAGEMENT ................ 75
Introduction ........................................................................ 75
The Economics of Broadband Wireless Access..................... 76
The Technology Revolution ................................................. 78
Adapting Spectrum Regulatory Models for BWA ................... 78
Defining Best Practices ....................................................... 83
Case Study: BWA Spectrum Allocation in Mauritius.............. 87
Conclusion ......................................................................... 90
6
6.1
6.2
6.3
6.4
6.5
6.6
VOIP AND REGULATION ................................................ 91
VoIP: Regulatory Evolution or Revolution? ............................ 91
The Pace of VoIP Market Development ................................. 95
Grappling with Change: Regulators’ Responses to VoIP ....... 96
Crafting New Regulatory Approaches to VoIP ...................... 100
End User and Consumer Issues ......................................... 105
Conclusion ....................................................................... 108
7
7.1
7.2
7.3
7.4
7.5
STEMMING THE INTERNATIONAL TIDE OF SPAM.. 111
The Spam Problem ............................................................ 111
An Outline of a Model Law ................................................. 115
Codes of Conduct ............................................................. 119
Education and Awareness ................................................. 122
Conclusion ....................................................................... 123
8
8.1
8.2
8.3
8.4
MAKING BROADBAND WORK FOR ALL .................... 127
What Is Broadband?........................................................... 127
Why Should I Care about Broadband?................................. 127
How Can I Get Broadband? ................................................ 128
What Can Regulators Do? .................................................. 128
REGULATORY TABLES .......................................................... 131
Table 1 – Countries with a separate Regulatory Authority .............133
Table 3 – Status of the main fixed-line operators .........................169
Table 4 – Level of competition ....................................................203
P
A
G
E
P
A
G
E
GLOSSARY OF TERMS .......................................................... 215
LIST OF TABLES..................................................................... 229
LIST OF FIGURES .................................................................. 228
LIST OF BOXES ...................................................................... 230
v
Trends in Telecommunication Reform 2006
1 MARKET AND REGULATORY TRENDS IN THE ICT SECTOR
In many countries, today’s broadband ‘missing link’ is the regulatory framework. Regulators
have before them an unprecedented opportunity to speed the uptake of broadband to enable the
Information Society.
1.1 What will it take to bring broadband to the
masses?
Bringing broadband to the masses is one of the major
challenges facing the global ICT community. Addressing this
challenge requires new thinking, and an end to business as
usual. The 7th edition of Trends in Telecommunication Reform is
designed to enable regulators and policy-makers to meet this
challenge.
What does an end to business as usual mean? Service providers are now offering the triple play of voice, internet and
broadcast as convergence moves from a dream to a part of
every day life, starting in the developed world and spreading
to the developing world. Music lovers download songs from
the internet onto their MP3 players, working parents order
their groceries online and come home at night to watch their
favorite TV shows broadcast over the internet, while international travelers book their flights and reserve their hotel rooms
on the web. Broadband is about more than travel and entertainment. International organizations like ITU use a broadband-enabled e-learning platform to provide capacity building
services to its membership and work with governments to
deploy broadband-enabled e-government services.
Wireless broadband technologies offer the prospect of
faster rollout of services, as well as portability and mobility.
Many broadband technologies can also be deployed incrementally, as demand develops, rather than requiring expensive network-wide upgrades. This means that a full range of players,
large and small, private and public, can harness the power of
these technological developments to become ICT service providers and close the broadband divide that exists between developing and developed countries, and between rural and urban
areas within countries. For example, already two developing
countries, Mauritius and TFYR Macedonia, have announced
plans to become entirely wireless broadband nations.
All of these market and technological developments are
exerting pressure on the current regulatory framework. How
will regulation change? Broadband regulation means a new
vision of reduced regulatory burdens, innovative incentives,
and coordinated efforts by all links in the broadband value
chain to unleash commercial deployment opportunities. Regulations can be carefully tailored to open the door to both large
and small-scale broadband providers. Broadband-promoting
C HAPTER 1
regulators can aim to make local communities and non-governmental organizations aware of the technologies and broadband
provisioning opportunities they could seize, and also coordinate
with other government and public institutions, such as universities, to drive demand for broadband-enabled health, education and government services. At the same time, regulators will
strive to revise outdated regulatory frameworks designed for an
earlier era. The new regulatory framework could be described
as a ‘less means more, old meets new’ approach. Less regulatory intervention means more business opportunities. Timetested regulatory principles such as transparency and open
competition will be applied to new technologies and the new
regulatory issues they raise. And the promotion of wireless
broadband technologies will require flexible and innovative
spectrum management practices.
Of course, broadband also poses new challenges. The
rapid rise of voice over IP (VoIP) – hastened by the spread of
broadband – is turning the old telecom business model on its
head. VoIP service providers have introduced a new business
model, providing voice services for free or bundled as a part
of a triple play package. What effect will this have on the business plans of traditional telecom operators? And what do these
developments mean for the current regulatory framework
designed for the old business model?
VoIP is not the only broadband-related challenge facing
the ICT sector. There is also the scourge of spam, which clogs
email inboxes and leads to internet fraud as well as poses network security problems such as the spread of internet viruses
and worms.
This year’s Trends in Telecommunication Reform contains
eight chapters addressing each of the broadband-related challenges and opportunities to enable regulators to harness the
potential of broadband to build a safe and secure Information
Society for all:
• This chapter provides an ICT market and regulatory overview to set the stage for the following chapters;
• Chapter Two defines what policy-makers and technologists mean by the term “broadband;”
• Chapter Three explores broadband technology at a more
technical level and looks ahead, providing a roadmap for
1
Figure 1.1: The Number of ICT Users Worldwide, 1994-2004
Fixed-line and mobile subscribers and internet users
Source:
ITU World Telecommunication Indicators Database.
regulators to plot the rapid technological changes that are
resulting in new opportunities and services;
•
Chapter Four examines the role of regulators in promoting broadband;
•
Chapter Five, recognizing the key role that wireless technologies are likely to play in the promotion of broadband
in developing countries, examines spectrum management
practices to promote broadband;
•
Chapter Six looks at current regulatory treatment of Voice
over IP (VoIP) and the regulatory road that lies ahead for
VoIP;
•
Chapter Seven addresses enforceable codes of conduct for
ISPs as a new legal tool regulators could deploy in their
fight against spam, and
•
Chapter Eight offers a conclusion and a look ahead.
1.2 ICT infrastructure deployment
The state of current ICT infrastructure deployment is
key to understanding the new technologies that can be used
to promote broadband access in developing countries.1 Today,
the majority of the world’s broadband subscribers are found in
developed countries, which have upgraded existing fixed line
telephone and cable TV infrastructure to provide broadband
services. Access to basic communications in the developing
world has largely been achieved through mobile communications. It is unlikely, therefore, that developing countries will
follow the same migration path to broadband services as developed countries. While developing countries will no doubt also
deploy some fixed line broadband services, broadband wireless
access is expected to play a key role for developing countries
seeking to foster the Information Society.
2
As shown in Figure 1.1, the growth of mobile lines continues to outpace the growth of fixed lines, and this is particularly evident in developing countries, a trend attributed to the
introduction of prepaid mobile services, rapid and cheaper
network deployment, a competitive environment as well as
the fact that mobile services provide access to a range of new
applications such as short and multimedia messaging services
(SMS and MMS). By the end of 2004, the world counted some
1.8 billion mobile subscribers (including both second and third
generation mobile subscribers), or 28 per cent of the world’s
population. Some 58 per cent of these mobile subscribers were
located in developing countries. Further information about 3G
subscribers is provided in section 1.2.2.
The number of fixed line subscribers worldwide, had
reached 1.2 billion lines, and a penetration rate of 19 per cent.
As shown in Figure 1.2, both fixed line and mobile penetration rates are lowest in Africa and Asia. Figures 1.2 also demonstrates that mobile teledensity rates outpace fixed teledensity in
every single region.
By year end 2004, there were an estimated 840 million
internet users in the world, representing 13.2 percent of the
total population.
1.2.1 Global Fixed Line Broadband Growth
As previously noted, the total number of internet users
(broadband and dial-up) around the world continues to
increase, having reached about 840 million at the end of 2004.
The total number of fixed-line broadband subscribers had
reached nearly 160 million at the end of the 2004. Broadband
internet subscribers represented approximately 2.5 percent of
the world’s population, and 38 per cent of all internet subscribers worldwide in 2004.
C HAPTER 1
Figure 1.2: Fixed vs. Mobile Teledensity by Region (per 100 subscribers)
Fixed telephone line penetration by region over the period of 1994-2004
Source:
Mobile cellular penetration, by region, 1999-2004
Figure 1.3: Broadband
Distribution of broadband subscribers by region, 2004
Source:
The vast majority of today’s broadband users are in the
developed world. Globally, Asia, Europe and the Americas
represent no less than 99 percent of all broadband subscribers,
the majority of which are in the wealthier countries of North
America, Western Europe and Asia (Figure 1.3). By contrast,
C HAPTER 1
Africa is home to only a fraction of broadband subscribers,
and many African countries have not yet launched high-speed
Internet services. Although Africa has the fewest numbers of
broadband subscribers globally, the number of subscribers
increased some thirty times in the two-year period from 2002
3
Figure 1.4: Global Distribution of Internet and Broadband Subscribers, 2004
Source:
ITU World Telecommunication Indicators Database, and OECD, the Organisation for Economic Cooperation and Development
Figure 1.5: Top 25 Broadband Subscribers, Non-OECD Countries (2004)
Source:
4
C HAPTER 1
Figure 1.6: Subscriber Growth in the Top 10 Broadband Countries (2000-2004)
Source:
to 2004, signally a healthy trend. A host of other developing
countries from other regions are also showing signs of vigorous broadband growth.
These healthy trends are reflected in Figure 1.4, showing
that approximately 25 per cent of all broadband subscribers
are in non-OECD countries, fuelled largely by China, which
has the second largest number of broadband subscribers in the
world, after the United States. Figure1.5 shows the top 25 nonOECD countries measured by number of broadband subscribers.
Indeed, attention should be focused on the considerable
new broadband deployment activity in nations throughout the
developing world, from the Arab States to Southeast Asia. For
example, TE Data and ISP Nile Online have rolled out DSL
internet lines. In Chile, carrier Telsur has initiated a broadband
development project, which has garnered a total investment of
USD 20 million in the last five years. Brazil has launched triple
play services including broadband, as has the Indian operator
MTNL.
Still, developing countries need to continue to grow their
broadband subscribers to compete with today’s top broadbandenabled economies. Figure 1.6 provides a snapshot of the top
ten countries in terms of broadband subscribers, and their
growth over the past four years.
Of the total broadband subscriber base, 32 per cent of
subscribers are using cable modems for access (51 million
subscribers), 62 per cent are using DSL (98 million subscribers), and 6 per cent (9.1 million subscribers) are using another
technology (for example, satellite, fibre-to-the-home and Ethernet LANs). Figure 1.7 depicts the breakout of broadband
access platforms by region. It is fair to assume that while there
may be other ways of achieving “broadband status” (for example, a VSAT connection), most broadband subscribers today
C HAPTER 1
are connected via DSL or cable modems. It is clear from the
data that DSL features most prominently in all regions of the
world except North America where cable modem technology
remains (at least for a now) more dominant. Meanwhile, the
majority of people – more than 60 per cent of all internet subscribers – still use dial-up connections.
1.2.2 Global Wireless Broadband Growth
IMT-2000 technologies, known popularly as 3G mobile,
are also starting to sprout broadband subscribers. By January
2005, 56 of the world’s economies were offering commercial 3G services and the total number of reported subscribers
accessing 3G technologies was 150 million, close to 60 per cent
growth from the previous year,2 and just shy of the 160 million fixed-line broadband subscribers. Of these 150 million
3G subscribers, 100 million are located in just three countries,
the United States (49.5 million), the Republic of Korea (27.5
million) and Japan (25.7 million). As described more fully in
Chapters 2 and 3, the dominant 3G technologies deployed to
date are W-CDMA and CDMA-1x. Figure 1.8 shows the top
ten 3G mobile markets worldwide in 2005, broken down by
these two standards.
As operators determine their migration preferences
toward the achievement of 3G networks, one open question is
whether the dominance of the GSM standard in 2G telephony
will continue to prevail in its more advanced 3G form, as some
industry groups predict. The breakdown of 2G standards use
by subscribers is highlighted in Figure 1.9.
Many countries are also looking at other wireless broadband technologies. Some countries are already extending
broadband connectivity through such BWA technologies, and
a couple of these have begun to do so on a nationwide basis.
(See Box 1.1).
5
Figure 1.7: Broadband Platform Distribution (by Region) 2004
Source:
1.2.3
The Poor Pay More for Less Broadband
As previously shown, broadband penetration is higher in
high-income countries. And yet, subscribers in low-income
countries pay more for less capacity. Given the fewer broadband users per capita in low-income countries, how does the
price and capacity of the service vary from country to country?
The average fee paid by subscribers in low-income countries
for broadband service is USD 291 per month, compared with
a mean price of USD 18 a month in upper-income countries.
While the poor pay more for their broadband service,
on average, they also receive less bandwidth. The average
high-speed downlink capacity in upper-income countries is
3.8 Mbit/s, compared with an average for low-income countries of 712 kbit/s.
In summary, while penetration is very small in lowincome countries, broadband subscribers there pay considerably more money for inferior service. Perhaps further research
and market studies could be undertaken to address whether
there are objective factors – such as the need to recover higher
network construction costs over a smaller initial subscriber
base – to explain why customers in lower-income and middleincome countries are paying more for less.
1.3 Market Trends in Privatizations, Investments and
Services
Privatization of state-owned incumbent operators often
affects the pace of regulatory reform, reducing or eliminating
concerns regulators and policy-makers may have about protecting legacy operators from the pressure of new market entrants
or new technologies that can provide lower cost and more
innovative services. Where privatizations have not occurred, or
where only partial privatizations have been implemented, there
6
is generally a greater tendency to balance the concerns of legacy
operators (for revenues and market share) with those of endusers (for low cost services). This delicate balance often results
in limitations on regulatory reforms. Privatization usually gives
regulators and policy-makers greater freedom to focus on the
interests of end users, and accelerate the reform process.
Although many governments now recognize the value
of privatizing their fixed line operators, privatization activity
has stalled over the past two years, with no new privatization
of state-owned fixed line operators either in 2004 or 2005, as
shown in Figure 1.10. A number of initial and further privatizations, however, are on the horizon, including in Tunisia and
Pakistan. Likewise, a number of companies are busy negotiating acquisitions or arranging financing to expand their network holdings in foreign markets. For example, one of China’s
leading mobile operators has held discussions concerning its
expansion into India; another mobile operator from Hong
Kong, China is planning expansion into India, Indonesia, and
Vietnam.
The ICT sector has also witnessed significant investment
in infrastructure equipment over the past year, signalling signs
of recovery following the market collapse in the early years of
the decade. According to some analysts, the total global market
in 2004 for wireless infrastructure was more than USD 39 billion, a 12 per cent increase in operator spending from 2003
levels.3 Expenditures on W-CDMA equipment likely crossed
the USD 10 billion threshold in 2005, largely based on spending generated by operators in China, Germany, Japan, Spain,
the United Kingdom and the United States.4 In fact, analysts
believe that China alone generated about 16 per cent of WCDMA-related spending in 2005.
The world was also witnessing the awesome growth of
companies benefiting directly from the growth of the internet.
C HAPTER 1
Box 1.1: Broadband Wireless Nations
A small set of countries has announced plans to vie (unofficially, of course) to become the world’s “first broadband wireless
nation.” Mauritius and TFYR Macedonia both are deploying fixed wireless broadband networks across the bulk of their countries,
using technologies that track the emerging 802.16 WiMAX standards.
In Mauritius, a small African island nation of 1.2 million people, the wireless broadband network reportedly already covers 60
per cent of the island and 70 per cent of the population. By the end of 2005, Mauritius intends coverage to reach a full 90 per cent
of the country.1
In TFYR Macedonia, a country of 2 million people, local network provider On.NET is deploying a broadband wireless network
across the country using Motorola’s Canopy radio system. The project is a unique partnership between the donor community, the
Macedonian Government, and the private sector. The Government of China has donated thousands of personal computers to be used
in primary and secondary schools. Complementing that donation, The U.S. Agency for International Development (USAID) is
providing broadband internet connectivity to 460 primary and secondary schools and 71 other sites through 2007. In October 2005,
the project partners announced an important milestone: 95 per cent of the country’s population was within reach of a broadband
wireless signal.
The project will guarantee a substantial countrywide customer base. This has created a business case for On.NET to make a
significant investment in a pervasive national wireless network. On.NET is free to sell capacity to additional corporate or consumer
subscribers throughout the country. Furthermore, in metropolitan areas, On.NET is deploying a mesh-based network to offer pervasive
hotspot connectivity in the country’s population centres.
1
This is being done using wireless technologies developed by Texas based Navini Networks.
Figure 1.8: Top 10 3G Mobile Markets Worldwide, 2005
Source:
ITU, The Internet of Things.
The 2005 December holiday season, for example, saw many
retailers celebrating record online sales volumes.
Traditional telecommunications and cable TV operators
are also benefiting from the rise of the internet and convergence. The past year has seen considerable emphasis on the
launch of triple play services combining video (television),
voice (telephone) and broadband (internet access) offerings
from the same provider. The triple play trend is most notable
C HAPTER 1
in the Americas and the Asia-Pacific region, and also in some
of the Arab States. Many telecommunication operators in
Brazil, for example, are investing in triple play packages.5 The
Uruguayan state-owned telecommunication company, Antel,
planned to launch triple play trials in 2006,6 while Chilean provider VTR-Metropolis invested in a plan to achieve nationwide
triple play coverage by 2010, serving 2 million homes.7 Service
launches were also set for June 2006 in Colombia.
7
Figure 1.9: 2G Standards Usage, by subscribers, world and by region (2004-5)
Today’s mobile subscribers are served by cellular
systems based on just a handful of major standards. The
two dominant standards are GSM and CDMA, with
GSM the most popular second-generation mobile standard in the world. The GSM Association estimated a
global user base of more than 1.5 billion as of the first
quarter of 2005--or three-quarters of the 2G subscriber
base. Approximately 31 million users are signing on
each month, with growth driven chiefly by developing
markets such as those in Africa, Latin America, Eastern
Europe and Asia. GSM is also gaining popularity in the
United States, reaching 55 million subscribers in 2005.
CDMA is the second most popular 2G standard, with the
CDMA Development Group reporting more than 270
million users. TDMA has a far smaller subscriber base,
and even this appears to be declining. Globally, reports
varied between 75 and 100 million TDMA subscribers,
with numbers in the United States declining some 23
per cent in 2004. The Personal Digital Communications (PDC) standard comprised the smallest segment
of global mobile users (chiefly dominating the Japanese
landscape). Many PDC users have been migrated over
to 3G services (through NTT DoCoMo) or had their
service terminated entirely, due to KDDI’s decision to
inaugurate CDMA 1X service.
Sources: GSM Association, CDMA Development Group, Mymobile.com News, NTT DoCoMo Corporate website, KDDI Corporate website, Vodafone Japan;
Source: The Diffusion Group Report, “U.S. Mobile Markets: Analysis & Forecasts.”
Notes:
8
TDMA subscriber numbers are included on the high end of analyst estimates.
C HAPTER 1
Figure 1.10: Privatizations, 1991-2005, World
Source:
ITU World Telecommunication Regulatory Database.
In Asia, for example, Pakistan Telecommunications Company Limited (PTCL) announced plans to launch a triple play
service, featuring telephony, cable TV and DSL services over
the same line.8 Nearby, in India, state-owned telephone company Mahanagar Telephone Nigam Ltd. plans to launch that
country’s first ‘triple play’ service, in Delhi and Mumbai; with
the state-owned sister company BSNL covering the rest of the
country.9 Bahrain was set to receive its first triple play voice,
video and data service from incumbent telephone company
Batelco, which struck a deal with an equipment supplier to
deploy infrastructure nationwide.10
In Eastern Europe, Romanian competitor Astral Telecom
completed a USD 1 million upgrade to its cable-based IP network to provide an enhanced triple play, with voice, video and
broadband data services.11 Spain’s dominant fixed-line operator,
Telefónica, launched a triple play package of voice telephony,
broadband internet and multi-channel TV services over traditional copper lines, reaching at least 100,000 subscribers.12 Italian fixed-line telecommunication operator Unidata launched
its new “ADSL2+” service, enabling users to download voice,
video and data at speeds of up to 24 megabits per second (Mbit/
s).13 Expanded service upgrades were also under way in France
and Germany.14
Mobile operators and equipment manufacturers are also
benefiting from convergence. One mobile operator recently
unveiled a personal digital assistant (PDA) that combines 3G
mobile technology with a miniature laptop personal computer
design and functionality, enabling mobile voice and internet
through 3G, Wi-Fi, and a Windows mobile operating system.15
C HAPTER 1
1.4 Competition
Over the past decade, the introduction of competition
into the second-generation mobile sector has been one of the
key factors linked to its success in increasing teledensity in
developing countries (along with prepaid cards and cheaper
network deployment costs). Many developing countries now
seek to replicate this success with broadband services. The
level of competition that is authorized by countries (based on
the number of competitive players), as well as the kinds of
services opened to competition, will remain key to these strategies. Competition in international services, in particular, the
international gateway, and leased lines, for example, is vital to
ensuring low cost internet access.
Figure 1.11 illustrates that competition is authorized
in basic telephone services (defined as local, long distance
and international) as well as in leased lines in more than 60
per cent of countries worldwide. Competition is authorized
in around 90 per cent of countries for services such as VSAT,
Cable TV, internet access, and IMT2000 (3G) offerings, and
by more than 80 per cent of countries for DSL and fixed wireless broadband services. The region with the lowest levels of
competition is the Arab States, where less than 40 per cent of
markets have been opened to competition. Europe is the most
competitive region, while Africa is nearly evenly split between
monopoly and competitive conditions. Competition prevails in
the Americas, Asia and, indeed, around the world as a whole
– by a ratio of 3 to 2.
1.4.1 Interconnection is Key to Competition
It is one thing for countries to authorize competition in
the provision of ICT services. It is another to ensure that new
market entrants are actually licensed or otherwise authorized
to provide services and are able to compete in the market on
9
Figure 1.11: Status of Competition Worldwide, 2005
Level of competition in selected services and networks, World
Level of competition in basic services, per region
Source:
a level playing field. One of the key issues in ensuring a level
playing field is a fair and transparent interconnection regulatory framework. Figure 1.12 illustrates the extent to which
interconnection agreements, or Reference Interconnection
Offers (RIOs), are made public, recognized as one of the best
practices in interconnection regulation today, ensuring that all
competitors are aware of and can benefit from the same interconnection rates. Transparency in interconnection agreements
is highest in the Americas, Asia-Pacific and Europe. In Europe,
10
pricing is public in approximately 72 per cent of countries,
but this is not the case in other regions. Worldwide, nearly
60 per cent of countries do not make interconnection agreements public, although about 58 per cent make pricing information available. This is one area regulators could assist, for
example, by publishing RIOs on their websites. ITU provides
links to published RIOs as part of the Regulatory Profiles section of its TREG website (http://www.itu.int/ITU-D/treg/profiles/Interconn.asp).
C HAPTER 1
Figure 1.12: Status of Public Interconnection Agreements and Pricing Information, 2005
Source:
1.4.2 Regulating Local Loop Unbundling
Opening up access to the local loop for competing operators is gaining ground as a means to foster full competition. At
the end of 2005, 72 countries worldwide had required unbundling (See Figure 1.13). The debates about unbundling mirror
discussions about competition in general. One argument is
that mandating unbundling creates disincentives for future
infrastructure investments. In the past, the local loop was
considered a “natural monopoly,” so it is not surprising that
many incumbents feel they should be able to retain the full
benefits of their investments. This argument is often coupled
C HAPTER 1
with the assertion that incumbents are in the best position to
channel their profits into large-scale expansion of last-mile
networks. Incumbents have not, however, always lived up to
those promises. Local-loop unbundling, therefore, is increasingly recognized as being important not only for competition
in traditional telephone services but also to prevent the incumbent’s monopoly from spilling over into the domain of broadband internet.16 Therefore, the goal of unbundling has often
been to make network components available to new market
entrants with greater commitment, expertise and incentives to
improve services.
11
Figure 1.13: Countries Requiring Local Loop Unbundling, 2005
Source:
1.4.3 Internet Exchange Points
Internet Exchange Points (IXPs), also known as network
access points (NAPs), are a way to maximize the existing infrastructural base for internet service provision in developing and
transitioning countries and foster competitive provision of
internet services. They allow Internet Service Providers (ISPs)
to exchange traffic between their networks through mutual
peering agreements at minimal cost, with better efficiency (in
terms of bandwidth and latency) and increased speed. Networks can interconnect directly, via the exchange, rather than
through third-party networks, and the cost of diverting traffic to upstream providers or to faraway places is minimized by
keeping it within and between adjacent ISPs. For many developing countries, IXPs keep traffic local or regional, instead
of routing internet traffic to hubs like New York or London
– just to reach an email account on a different ISP in the same
country or to exchange internet traffic with a neighbouring
country. Figure 1.14 illustrates the number of countries, per
region, that participate in national or subregional IXPs.
A study prepared for the 2004 Global Symposium for Regulators presented several issues related to creating a conducive
environment for IXPs, including the importance of clearing
the regulatory obstacles that exist at subregional levels, ensuring competition at the level of the international gateway, opening up VSAT use, allowing for co-location of equipment in an
incumbent operator’s facilities, and ensuring that licensing or
other additional burdens are minimized or eliminated.17
1.5 National Regulatory Authorities, Who Rules?
The establishment of a separate regulator is one of the
most visible signs of sector reform. Separate regulatory agencies cannot compensate for boom and bust cycles in financial
markets and the macro-economy, but they can certainly go a
12
long way in laying the grounds for a favourable investment climate and promoting market opportunities.18 It is well documented that the degree of actual and perceived autonomy from
government control and industry influence is a key indicator of
the effectiveness of a regulator. Regulatory procedures should
be clear, transparent, and predictable. Regulators, in enforcing these procedures, should be accountable, and should have
sufficient credibility and authority to enforce the relevant laws
and regulations.
Over the last five years, the number of regulatory authorities worldwide has increased by approximately 36 per cent,
with new regulators on the verge in Samoa and Liberia. Regulatory activity is most notably on the rise in the Arab States,
Africa and Asia. Figure 1.15 depicts the growth of these regulators worldwide.
Meanwhile, figure 1.16 shows the regional variation in
countries that have functional regulators in place, breaking out
the percentage of countries in each region that have established
regulatory authorities. It is evident that the presence of regulators in the Americas and in Africa (relative to the total number
of countries in each region) is higher than in the Arab States
or in Asia. In absolute numbers, the region with the highest
number of regulatory authorities is Europe.
Merely establishing a regulator – while a step in the right
direction – does not ensure effectiveness or relevance. Priorities have to be established, the necessary statutory framework
in which the regulator will operate must be created, procedures put in place, and relationships between stakeholders
established. The way in which these processes work together
and are respected within a general context of good governance
will determine the effectiveness of the regulator and ultimately
the success of the market.19
C HAPTER 1
Figure 1.14: Number of Countries with National and SubRegional IXPs, by region, 2005
Source:
Figure 1.15: Growth of Regulators Worldwide, 2005
Source:
Some countries possess more collegial, cooperative and
intertwined policy formulation climates than others. Some
governments permit informal decision-making processes,
while others mandate use of formal mechanisms. Retroactive
applications of influence can reverse or undermine regulatory
decisions. More fundamentally, some countries may consider
their regulatory bodies independent and separate, even when
in reality they may be subordinate to a ministry or subject to
C HAPTER 1
oversight from a higher authority. In the end, the most important element is effectiveness.20
Another key question facing policy-makers and regulators is what kind of regulatory body to create. Worldwide, most
countries have what can be termed a “single-sector” regulatory
authority for telecommunications. However, with the convergence of different communication services and technologies
“converged” regulatory agencies have started to emerge. Countries such as Austria, Finland, France, Italy, the Netherlands,
13
Figure 1.16: Separate Regulators, by Region, 2005
Source:
Saudi Arabia, Singapore, South Africa and the United Kingdom have taken this path.21 There is evidence that a unitary,
converged regulatory structure may yield more efficient internal administration, institutional flexibility and lower regulatory
costs.
Interestingly, multi-sector regulatory authorities have also
emerged in the last few years, giving oversight not only of the
telecommunication sector, but also of other industry sectors
with common economic and legal characteristics (for example,
water, energy or transportation). Costa Rica, Gambia, Germany,
Jamaica, Latvia, Luxembourg, Niger and Panama have chosen
this model, which has long been standard for public utility
commissions in individual states in the United States.22 One of
the arguments for such a structure is centralizing and optimizing limited regulatory skills, as well as reducing costs through
economies of scale. However, there are know disadvantages of
this model including dilution of sector specific expertise, failures of a multi-sector authority cascading through multiple
regulated sectors, increased risk of political capture, delays in
reform, etc.
The fourth approach is when countries have opted to have
no national regulatory authority per se, and the functions of
sector regulation rely on competition and antitrust rules.
1.6 Regulatory Challenges of VoIP
Many of the world’s carriers have been persuaded to
deploy IP-based networks that can carry both voice and data.
In this way, operators are able to invest in a single network that
can be used more efficiently for many different forms of traffic. Many of these operators have started to offer VoIP to their
customers.
Recognition of this shift in practice and service is widespread and international. In fact, international VoIP increased
by 35 per cent from 2003 to 2004 (See figure 1.17). For example, 20-25 per cent of all historic operators in Africa were
using VoIP to carry part of their international traffic in 2004.23
14
Telkom Kenya currently offers a VoIP-based international service. Five African carriers – in South Africa, Botswana, Angola,
Namibia and Uganda – have announced that the introduction
of IP-based networks is imminent, while Mexican incumbent
Telmex (amongst various others carriers) has already implemented IP for the majority of its core network. Among the
Arab States, Oman’s incumbent Omantel has committed itself
to creating an end-to-end IP communication services network.
VoIP is also being offered by new market players, which
are often viewed as a threat to traditional PSTN operators. VoIP
may be offered by ISPs, in internet cafés and, more recently, by
companies from abroad that have no local presence, yet whose
impact on local market conditions is felt quite strongly. The
impact of VoIP is felt from the loss of outgoing international
retail traffic (as customers search for the lowest cost international rates) to reductions of incoming international settlement
traffic (as traffic from VoIP customers abroad skirts the international settlement rate system).
Regulators participating in the 2005 ITU Global Symposium for Regulators (GSR) recognized that although VoIP
poses increasing challenges to legacy operators, it also brings
new opportunities to end users for more affordable services.24
In many ways, the rise of VoIP has crystallized the delicate balancing act that many regulators have been performing as regulatory reform has been implemented ever more widely. Weighing
in on one side of the scale are the commitments of the World
Summit on the Information Society (WSIS) to encourage low
cost access to ICT services, while the other end of the scale
balances the desire to protect incumbent operators-especially
when incumbents remain at least partially government owned.
It is not surprising, therefore, that the rise of VoIP has
prompted an array of regulatory responses, from outright bans
to full legalization. The question of whether to allow or to
prohibit VoIP, however, is only one of many issues prompted
by the rise of VoIP. Other issues include developing regulatory
frameworks for the interconnection of circuit-switched and IPbased networks as well as transitioning to interconnection in
C HAPTER 1
Table 1.1: Newly Created Regulatory Authorities, 2005
Regulator
Year created
Structure
Cape Verde
Country
Instituto das
Comunicações e
das Téconologias de
Informação
2004
Collegial body of 3
members
Functions
Trinidad & Tobago
Telecommunications
Authority of Trinidad
and Tobago
2004
Collegial body of 11 Interconnection, price regulation, Radio
members
frequency allocation, numbering, monitor
service quality, quality of service
standards, and licensing and universal
service with the sector Ministry
Thailand
National
Telecommunications
Commission (NTC)
2004
Collegial body of 7
members reporting
to House Senate
Interconnection, price regulation,
technical standards, radio frequency
allocation, numbering, type approval,
monitor service quality, quality of service
Qatar
Supreme Council for
Communication & IT
2004
Headed by
Secretary General
Mandate to create a legal and regulatory
environment that promotes development
of ICTs
United Arab
Emirates
Telecommunications
Regulatory Authority
2004
Collegial body
of 5 members
reporting to UAE
Telecom Supreme
Committee
Licensing, interconnection, price
regulation, technical standards, radio
frequency allocation, numbering, type
approval, monitor service quality,
universal service, quality of service
standards
Niger
Autorité de
Régulation
Multisectorielle
(ARM)
2004
Headed by a
president with a
sector director
for each sector it
regulates.
Transport, energy, water, and
telecommunication
Iran
Communications
Law passed
in 2003 and
operational
in 2005
Collegial body of 7
members reporting
to Minister of ICT
Licensing, interconnection, technical
standards, radio frequency allocation,
numbering, type approval, monitor
service quality, universal service, quality
of service standards
Afghanistan
Afghanistan Telecom
(ATRA)*
Law passed
in December
2005.
Operational
mid-Jan
2006
TFYR Macedonia
Agency for electronic
communications
2005
Licensing, interconnection, price
regulation, technical standards, radio
frequency allocation, numbering, type
approval, monitor service quality,
universal service, quality of service
All regulatory functions, including
licensing & compliance, spectrum
planning & assignment, numbering,
ensuring network interconnection,
promoting competition and consumer
protection
Collegial body of 5
members reporting
to Parliament
Licensing, interconnection, price regulation, technical standards, radio frequency
allocation, numbering, type approval,
monitor service quality, universal service,
quality of service standards
* The ATRA was established from the merger of the Telecom Regulatory Board, created in 2003, with the State Radio Inspection Department. See
http://trb.gov.af/trb.htm
Source:
a fully IP-based world. There is also a range of issues arising
from the fact that VoIP customers can use the same VoIP service nomadically, rather than from one fixed line location or one
mobile terminal. The nomadic nature of VoIP is prompting
challenges to traditional practices in numbering and emergency
services. The variety and intricacies of the regulatory treatment
of VoIP are explored in Chapter 6.
C HAPTER 1
The rise of VoIP is emblematic of the crossroads at which
the ICT sector now finds itself. Regulatory practices and
wisdom built upon the experiences of the heavily regulated
PSTN era are now meeting head on the largely unregulated
internet world. Which model will apply as these two worlds
converge? Or will hybrid or entirely new regulatory models be
developed? Chapter 6 begins to identify the issues and prac-
15
Figure 1.17: Growth of International Traffic
Source:
ITU World Telecommunication Indicators Database and TeleGeography Global Traffic Statistics 2006 report. Telegeography Research is now part of
PriMetrica, Inc. (see www.primetrica.com)
tices facing regulators today and in the years ahead as IP networks replace circuit switched infrastructure.
structure), developing countries in Africa, the Americas and
Asia appear less eager to regulate WLANs than their counterparts in Europe, where WLANs are just one potential solution
to broadband connectivity.
1.7 Spectrum Management
A pragmatic framework for managing spectrum resources
effectively includes ways of facilitating deployment of innovative broadband technologies. Spectrum management to promote broadband access is discussed more fully in Chapter Five.
Strategically balancing the use of unlicensed and licensed spectrum, for example, is becoming a key component of effective
regulation.
Across regions, a lack of consistency among national spectrum policies – particularly on unlicensed “commons” models
– is becoming an issue. Lack of coherence from country to
country in the fees and costs associated with spectrum access
may be manageable in the short-term, but it is likely to be fundamentally problematic in the long term, as innovations sweep
through the market in disregard of state borders. Gradual convergence of regional policies and fees has been encouraged,
particularly through actions of the European Commission, for
example.
The disparity between massive fees paid for 3G licences
(in Europe) and the unlicensed use of wireless local area networks (WLANs) looms as a potential regulatory issue. It will
loom larger if both technologies converge into a single broadband wireless market. For now, it appears that licensing as a
regulatory mechanism is being applied in limited ways to
the realm of WLANs. Figure 1.18 shows that the majority of
countries in all regions do not to require spectrum licences for
WLANs. Perhaps in light of the numerous, uncontrolled and
rather random deployments of WLANs as part of small-scale
development projects (often launched in lieu of fixed-line infra-
16
1.8 Dealing with Spam
Over the last decade, the unbridled growth of spam has
gained increasing attention, not only due to its inconvenience
and cost, but perhaps even more importantly, because spam
often carries viruses and worms or poses other network security issues, or is used a vehicle for fraudulent behaviour. Today,
there is general agreement about spam’s core characteristics,
including that it consists of unsolicited electronic messages
sent in bulk. “Spammers now employ a variety of advanced
upload methods such as open mail relays, insecure Web
proxies, malformed CGI scripts and zombied clueless-user
machines.”25 This means that spam messages tend to be identical and are sent indiscriminately to selected recipients. Most
experts involved in the fight against spam counsel in favour of
a multi-pronged approach, including technical solutions, legal
and regulatory actions, end-user education and international
cooperation.
According to some analysts, spam accounted for around 70
per cent of all e-mail traffic by mid-2005 (see Figure 1.19). The
costs associated with spam are difficult to determine, although
it is logical to assume that it puts pressure on ISPs in terms of
reduced bandwidth and increased storage costs – not to mention the burden of dealing with customer complaints. The
European Commission has estimated the cost of spam to internet users worldwide to be around EUR 10 billion a year, and a
recent study has estimated the loss of productivity due to spam
messages to be at USD 1,930 annually for each employee.26 In
C HAPTER 1
Figure 1.18: Are Licences Required for WLANS, 2005?
Source:
Figure 1.19: Spam as Percentage of Emails Worldwide, 2003-05
Source:
Message labs.
marked contrast, the costs of startup and operation for spammers are extremely low, and the architecture, based on Simple
Mail Transfer Protocol (SMPT), allows them to work anonymously.
The success of legislating and making policies effective in
countering spam has been limited thus far. In 66 per cent of all
countries, there is no single, identifiable entity responsible for
combating spam (see Figure 1.20). Only thirty-two countries
have passed anti-spam legislation. As a region, Europe has the
greatest focus on anti-spam measures, although international
C HAPTER 1
attempts at standardizing business practices – or at least harmonizing ISPs’ approaches in countering spam – are growing.
To date, anti-spam laws have focused mainly on tracking
down and prosecuting spammers. Such anti-spam laws require
considerable investigative and enforcement resources, the very
resources that often are in short supply in developing countries.
While anti-spam laws targeted at spammers remain an essential
tool in the anti-spam arsenal, their use by developing countries may more likely be as the foundation for international
cooperation. Anti-spam authorities with more experience and
17
Figure 1.20: Spam Regulation, 2005
Source:
resources may seek to work with regulators in developing
countries in tracking down and prosecuting spammers. Having
an enforceable anti-spam law in place as part of a coordinated
international effort will facilitate action against spammers
acting (and hiding) across multiple jurisdictions.
But the time may also be ripe for anti-spam authorities to
expand their efforts to include working with ISPs, who can be
instrumental in fighting spam. Chapter 7 therefore looks not
only at the components of anti-spam laws targeted at spammers, but proposes the establishment of enforceable codes
of conduct to be developed by ISPs, and then approved and
enforced by regulators. Such a system of ‘managed self-regulation’ would require ISPs to prohibit their customers from
using that ISP as a source for spamming and related bad acts,
such as spoofing and phishing, and not to enter into peering
arrangements with ISPs that do not uphold similar codes of
conduct. Rather than continue to rely upon chasing individual
spammers, regulators in the most resource-constrained countries in particular would be more likely to succeed by working
with and through the ISPs that are closer to the source of the
problem, to their customers, and to the technology in question. The regulator’s job would be to ensure that ISPs within
their jurisdiction adopt adequate codes of conduct and then to
enforce adherence to those codes.
While some ISPs can be expected to resist even such lighthanded regulation, the advantage is that it places all ISPs on a
level playing field. Under current practices, responsible ISPs
find themselves bearing the brunt of the costs of spam. This
explains why some ISPs have begun suing spammers for damages, an option that may not be available in all jurisdictions.
The goal of managed self-regulation is to reduce spam in a way
that protects responsible ISPs. ISPs that implement responsible, effective anti-spam measures should be rewarded for their
18
good behaviour. One means of rewarding those responsible
ISPs is for regulators to hold their irresponsible competitors
accountable. Regulators can also make consumers aware of the
good works of the best ISPs, for example, by certifying ISPs
that enforce their codes of conduct and allowing such ISPs
to use the regulator certification in their advertising. As with
many other telecommunication-related policy issue that is
salient across national borders, the importance of consistency,
shared strategic approaches and international cooperation is
paramount.
1.9 Conclusion
Since the release of the 2004 edition of Trends in Telecommunication Reform, the global ICT sector has retained its
dynamism, although potential challenges to continued growth
loom large ahead. The prevailing trend of liberalized markets,
privatized operators, separate regulatory authorities and new
regulations aimed at tackling the phenomena of convergence
are very real. It is unlikely that countries today will attempt
to shut down competition and bolster the formation of new
monopolies, although some operators upgrading to next generation networks will have to be watched closely to counter
such tendencies. Regulators will continue to play a key role
in ensuring that the values of transparency, cooperation, and
market competition now being championed in the ICT sector
become more real and apparent over time.
As institutions and frameworks are gradually being redesigned, a clear message is being disseminated: the role of regulators is a critical catalyst to the process of reform in the ICT
sector. Broadband internet access (whether through fixed lines
or wireless) is becoming increasingly relevant to the demands
of subscribers in developed and developing countries alike.
C HAPTER 1
The essence of voice telephony is being transformed. Nextgeneration networks are being designed and developed, even as
3G services begin to gain widespread acceptance. Through all
of these trends, one thing appears certain: the sector is tending
toward a more open, competitive, and transparent model, in
which governments, operators, development agencies, educational institutions, civil society groups, and end users all have
equally important stakes.
1
A comprehensive picture of the state of current ICT infrastructure deployment may be found in the 2006 World Telecommunication Development Report, being released
at the same time as this publication.
2 3G Today Newsletter. January 2005. Volume 2. Issue 1. See: http://www.3gtoday.com/wps/portal/!ut/p/kcxml/04_Sj9SPykssy0xPLMnMz0vM0Y_QjzKLN4r3DAbJmMUbxBub6keiijjCBXw98nNT9b31A_QLckMjyh0VFQEBN7t9/delta/base64xml/L3dJdyEvUUd3QndNQSEvNElVRS82XzJfSjI!?newsletterId=1180
3 Aytar, Ozgur, Pyramid Research cited in CommsDay Global, November 5, 2004, p.4.
4 “Pain or Gain: The Year Ahead for Mobile”, 3G Analyst Predictions Lucent Newsletter, Issue 18, February 2005.
5 “Brazil Telecom Launches Triple Play”, Telegeography Commsupdate, October 14, 2005.
6 “Antel Trials 3G and Triple play Services”, Telegeography Commsupdate, September 28, 2005. Link: http://www.telegeography.com/cu/article.php?article_id=9228.
7 “Chilean companies invest heavily in triple-play networks”, Telegeography Commsupdate, September 13, 2005. Link: http://www.telegeography.com/cu/article.
php?article_id=9004.
8 “PTCL To Launch Triple Play Service”, Telegeography Commsupdate, May 3, 2005. Link: http://www.telegeography.com/cu/article.php?article_id=6995.
9 “MTNL Launches triple-play (India)”, Telecomfinance.com, September 2005. Link: http://www.telecomfinance.com/nodes/Main-Pages.html.
10 “Batelco to Launch Triple Play”, Telegeography Commsupdate, June 20, 2005. Link: http://www.telegeography.com/cu/article.php?article_id=7690.
11 “Astral Telecom carries out USD1 million voice-over-cable upgrade”, Telegeography Commsupdate, September 21, 2005. Link: http://www.telegeography.com/cu/
article.php?article_id=9116.
12 “Telefónica talks up triple play”, Telegeography Commsupdate, October 14, 2005. Link: http://www.telegeography.com/cu/article.php?article_id=9465.
13 “When in Rome… download at 24Mbit/s”, Telegeography Commsupdate, June 22, 2005. Link: http://www.telegeography.com/cu/article.php?article_id=7736.
14 “Swisscom runs into May triple play delay”, Telegeography Commsupdate, May 31, 2005. Link: http://www.telegeography.com/cu/article.php?article_id=7433.
15 “Orange Launches Laptop Style 3G Phone/PDA For Business”, Businessmobile.com, September 22, 2005, Link: http://www.businessmobile.com/index.php?name
=News&file=article&sid=51.
16 Regulators from West Africa recognized the importance of local loop unbundling in a set of guidelines on interconnection agreed by the West African Telecommunications Regulators Assembly (WATRA) http://www.itu.int/newsarchive/press_releases/2005/12.html
17 Russell Southwood, “Via Africa: Creating local and regional IXPs to save money and bandwidth”, International Telecommunication Union, 2005, p. 23, Link: http://
www.itu.int/ITU-D/treg/publications/AfricaIXPRep.pdf.
18 Melody, W.H. “Stimulating Investment in Network Development: Roles for Telecom Regulation”, World Dialogue on Regulation of Network Economies, March
2003, p. 9.
19 Forthcoming ITU/infoDev ICT regulation toolkit, Module on legal and institutional aspects of regulation, 2006. Link: http://www.ictregulationtoolkit.org/
20 For further reading on effectiveness, consult the ITU 2002 Trends in Telecommunication Reform: Licensing in an Era of Convergence. http://www.itu.int/ITU-D/
treg/publications/Trends02_ExecSummary.pdf
21 Forthcoming ITU InfoDev Toolkit, Organizational and Institutional Approaches to Regulation, Chapter 6, p. 9-10.
22 See http://www.itu.int/ITU-D/treg/index.html. See ARESEP (Costa Rica) http://www.aresep.go.cr/cgi-bin/menu.fwx; OUR (Jamaica) http://www.our.org.jm/; Ente
Regulador de los Servicios Publicos (Panama) http://www.ersp.gob.pa/default.asp. inks to U.S. State PUC’s can be found at http://www.dps.state.ny.us/stateweb.
htm.
23 Muleta, John. VoIP Chapter X, pp.
24 See http://www.itu.int/ITU-D/treg/Events/Seminars/2005/GSR05/Documents/chairmansreport.pdf
25 Race, Jeffrey, “You needn’t eat spam (or worms) The real reasons why spam still exists today – and what to do about it”. Free Software Magazine, Issue 6, August
2005.
26 Nucleus Research “Spam: The Serial ROI Killer”. Link: http://www.nucleusresearch.com/index.html
C HAPTER 1
19
2 WHAT IS “BROADBAND”?
Authors: Michael Best, Georgia Tech; and Bjorn Pehrson, Swedish Royal Institute of Technology
There is a lot of discussion around the world about the
advent of broadband technologies and services – and justifiably
so. The long-awaited broadband revolution now appears to be
gaining momentum, in developing countries as well as developed ones. This chapter seeks to take the measurements of this
broadband growth by defining what policy-makers and technologists mean when they say “broadband. This will set the
stage for Chapter 3, which explores broadband technology at
a more technical level. Chapter 3 also looks ahead, providing a
roadmap for regulators to plot the rapid technological changes
that are resulting in new opportunities and services, on a nearly
daily basis.
The range of broadband systems that are in the market
and commonly employed today can be organized into three
broad families of technologies:
• Broadband wire-line networks, including DSL over
twisted-pair copper cables, cable modem services over
cable TV systems, and fibre networks;
• Broadband wireless solutions, including 3G mobile
services, wireless LANs (WLANS) and other fixed and
mobile wireless access solutions; and
• Non-terrestrial options, including satellite systems
employing very-small aperture terminal (VSAT) dishes.
This chapter will discuss the types of equipment, infrastructure and software that are needed to deploy each and their
viability for rural and underserved areas of developing countries. The picture that emerges is that broadband development
is no longer the exclusive preserve of developed countries and
high-income communities. Increasingly, developing countries
are employing low-cost technologies – many of them wireless
– to strategically introduce broadband capabilities.
Moreover, it is not just large, incumbent operators that
have opportunities to deploy broadband facilities and services.
There are opportunities for non-traditional operators – such as
universities, local governments and community-based groups
– to invest in broadband capabilities and link their systems to
the broader internet. The combination of non-traditional
entrepreneurs, local governments and major incumbents may
provide the formula to drive broadband development in many
countries.
C HAPTER 2
2.1 What Do We Mean When We Say “Broadband”?
Like many terms used in today’s fast-moving technology
sector, the term broadband is not well defined. The word was
originally used in the network engineering community to signify transmissions carrying multiple channels simultaneously.
This was contrasted with baseband, which involved transmitting on only a single channel at any one time. Today, however,
“broadband” is used much more frequently to indicate some
form of high-speed internet access or data transmission.
Deciding which networks provide sufficient capacity to
be called broadband is open to debate. There have been many
attempts to associate the term with a particular speed or set of
services, but in reality, broadband is a moving target. Internet
speeds are increasing constantly, and with each new advance,
marketers eagerly emphasize just how blazingly fast the latest
connection speeds are.
The speed of a network is usually expressed in terms of its
data transmission rate, which is measured in kilobits or megabits per second (this is also known as the bandwidth of the network).
What is not up for debate is that today’s dial-up internet
access speeds, topping out at about 56 kilobits per second (kbit/
s), are not broadband connections. But beyond that, defining
broadband seems to be subjective. The U.S. Federal Communication Commission (FCC) has defined broadband as starting at 200 kbit/s. The OECD sets the bar at 256 kbit/s, and the
ITU defines broadband as a combined capacity (upstream and
downstream) totalling 256 kbit/s or more.
A good example of the broad range of definitions can be
found within a single country. The Swedish IT Commission
(1994-2004) has defined broadband as supporting a formidable 5 megabits per second (Mbit/s) upstream and downstream.
But the Swedish government has said that broadband starts at
2 Mbit/s, up and down, while Sweden’s incumbent operator,
Telia, defines it as at least 500 kbit/s up and down. In Swedish
metropolitan area networks (MANs), meanwhile, 10 Mbit/s up
to 100 Mbit/s has become the standard. While there may be
a diverse array of parameters in the marketplace, for the purposes of this report, the term broadband refers to data rates that
correspond to the ITU’s definition in this section.
21
Beyond precise data-rate thresholds, perhaps a more
useful way to define broadband is by discussing what one can
do with it. This certainly includes fast internet downloads
from the Web. But it also should include compact disc-quality
streaming audio, fully interactive voice services such as VoIP,
some level of interactive video chat services (if not full-capability video conferencing), and at least reasonable-quality streaming video services (if not full, DVD-quality video on demand).
Note that receiving this wish list of services and applications is
not simply a matter of available bandwidth. Interactive applications such as VoIP also require little or no latency (delays), low
error rates, and minimal jitter (the result of data arriving at its
destination out of order).
2.2 A Short Broadband Taxonomy
As noted in the introduction to this chapter, broadband
networks can be divided into three major categories: broadband wire-line networks, broadband wireless networks, and
non-terrestrial (that is, satellite) networks.1 These categories
reflect the different development paths for broadband technologies that are now available in the marketplace. The following
subsections trace those different paths and briefly describe the
current state of development of standards and capabilities.
2.2.1 Broadband Wire-Line Networks
In general, internet access obtained its first foothold as an
outgrowth of wire-line networks. The earliest internet access
was, of course, through the public switched telephone network
(PSTN), via dial-up modems or leased telephone lines, to an
internet service provider (ISP). Data communications using a
dial-up telephone connection require an analogue modem at
both ends of the telephone line. Traditional modems encode
data in the same frequency band as a voice call (up to 4 kilohertz). The user can either speak or send data. Data rates vary
between 2.4 kbit/s and 56 kbit/s, depending on the quality of
the analogue copper telephone line, whether or not the network operator’s central office switch is digital, whether the
switches are clock-synchronized, and whether the switches are
connected using modern links, such as fibre or microwave.
From an economic point of view, dial-up modems are
cheap and typically integrated into personal computers. Carrierclass dial-up servers required at the ISP end are quite expensive,
while lower-class equipment that can be used by rural entrepreneurs on a small scale is cheap (less than USD 500) and
supported by open-source software. From the end user’s point
of view, the cost typically includes the local telephone call tariff,
plus the ISP’s internet access fee, which may be flat-rated or
connection-time based (in many areas of the United States, the
local calling rate is also flat). Dial-up service continues to be
a standard form of internet access in many countries, but, as
previously mentioned, it does not meet most definitions of
broadband. Dial-up service has paved the way, however, for the
broadband technologies discussed in the following subsections.
22
2.2.1.1
Digital Subscriber Line (DSL)
DSL is actually a family of technologies that provide a digital connection in an unused part of the frequency spectrum
of the telephone network’s copper-wire subscriber line.2 The
circuit-switched voice and packet-switched data connections
can thus be used independently of each other. DSL technology provides a significant enhancement to the already installed
PSTN base, protecting the sunken investment value of the
copper network. For that reason, DSL is one of the most popular broadband technologies. According to the ITU’s World Telecommunication Indicators database, DSL is used to provide over
60 per cent of worldwide home broadband connectivity.
The bandwidth that DSL systems can provide has been
increasing. There are now systems that can provide transmission speeds from 256 kbit/s up to 1.2 Mbit/s upstream and 512
kbit/s to 28 Mbit/s downstream. DSL bandwidth limitations are
caused by the attenuation of signals at higher frequencies. The
amount of attenuation on any network depends on the quality of the copper lines and their installation. Moreover, DSL’s
viability in any given location depends on the distance between
the subscriber and the exchange, which usually must be within
5 kilometres.
To deploy DSL, equipment must be added at both ends of
the subscriber line. At the user end, a DSL modem and a cheap
passive splitter must be installed. The passive splitter plugs into
the existing telephone socket and splits the incoming signal
between the telephone and the DSL modem. On the other side
of the modem, the user can directly connect a computer or set
up a LAN via a customer premises gateway (often for less than
USD 100). The DSL modem converts signals from the data
format used in the local area network environment (mostly an
Ethernet LAN or IEEE802.3 signal), into a digital audio stream.
On the operator side, before the subscriber line is connected
to the telephone exchange, the DSL circuits are separated and
terminated in a digital subscriber line access multiplexer (DSLAM),
which aggregates the digital connections from different users
and feeds them to the ISP network.
2.2.1.2
Cable TV System Networks
Cable TV (CATV) networks use coaxial cable to reach
all users in a point-to-multipoint topology. Initially, of course,
CATV was employed only for distribution of television channels in a tree-structured network, created by using passive
splitters. Broadband communication over cable TV networks
is accomplished by transferring data via unused bandwidth in
the cable, in a way similar to what DSL does over the PSTN.
The standard is Data over Cable Service Interface Specification
(DOCSIS).3 The basic data rates are 54 Mbit/s downstream
and 3 Mbit/s upstream. An ISP connects to the cable company’s central office (known as the head end by CATV operators)
and uses the cable network to connect to users.
As with DSL, equipment has to be installed at both the
head end and at the customer premises. On the operator’s
side, a cable modem termination system (CMTS) is installed at the
head end. It separates the digital communication channel from
the television circuits, aggregates connections from different
users and feeds them into the ISP network. On the user’s side,
C HAPTER 2
a splitter and a cable modem must be installed at the home
(most cable modem subscribers are residential). The splitter
divides the incoming signal between the TV set and the cable
modem. On the other side of the modem, the user can connect a computer or a residential gateway via an Ethernet port
and a USB telephone, allowing VoIP service if provided by the
ISP or other provider.
environment (EDGE). GPRS offers maximum speeds of 171.2
kbit/s, while EDGE can triple those rates. A typical cell radius
for these 2.5G networks is 500-1 000 meters, meaning that to
achieve complete coverage, every point in the coverage area
can be no more than 1 km away from an antenna. GPRS and
EDGE technologies allow mobility at vehicular speeds and can
handle seamless handoffs between cells.
From an economic point of view, broadband over cable
is favourable in areas where there is already an existing cable
network. The tree-structured, point to multipoint technique
has, however, several severe disadvantages compared to pointto-point solutions:
• For the regulator, it is less attractive, since it may prevent
local loop unbundling on some technical layers of the network;
• For the operator, it is more complex to plan, manage and
upgrade; and
• For the user, the performance depends on how much traffic there is from other users connected to the local transmission lines.
Meanwhile, upgrades from 2G to 2.5G networks can
be expensive, since they generally require both software and
hardware changes at the base station. They may also require
improvements to backhaul networks, as well as increased connectivity at the core network. In addition, in many settings,
high costs for spectrum licensing (for example, in the 2 GHz
band) add to the upgrade costs. Subscriber handsets also need
to be upgraded to support 2.5G, and handset replacement
cycles have often been a central component of adopting new
technologies. While estimating the cost is difficult, common
implementations have experienced costs per base station above
USD 100,000. In other words, upgrade costs can be on the
order of USD 50 or more, per subscriber, in low-population
density areas.
Users in a neighbourhood (typically, 100 to 2000 homes)
share the available bandwidth provided by a single coaxial cable
line. Therefore, connection speeds can vary between 10 Mbit/s
and a few kbit/s, depending on the volume of traffic from other
users. While most networks share a fixed amount of bandwidth
between users, cable networks are generally spread over larger
areas and require more attention to performance issues. The
broadcasting technique also raises concerns regarding security
and privacy. To address these concerns, the DOCSIS standard
includes encryption and other privacy features that are supported by most cable modems.
2.2.2 Broadband Wireless Access
The ITU defines broadband wireless access (BWA) as
encompassing either mobile or fixed access technologies
that provide connections at speeds higher than the primary rate (for example, 2 Mbit/s). Up until now, wireless
broadband has somewhat lagged behind the development curve
of its wire-line cousins. Throughout the 1990s, mobile networks, for example, were primarily viewed as voice offerings,
and broadband fixed wireless systems (many of which relied
on line-of-sight technologies) did not enjoy the same take-up
rates in developed countries that DSL and cable modem services did. Wireless broadband options are, however, becoming
increasingly more available and more functional (see Chapter
3). The following subsections trace the development paths of
mobile and fixed wireless terrestrial systems.
2.2.2.1
Mobile Services: GSM and CDMA
Second generation (2G) GSM networks were initially
implemented in Europe and then Asia. They have since
been installed across much of the world. The original GSM
systems support only very limited data capacity (well below
that required to be labelled broadband). On the path toward
broadband, the next step for GSM-based data services has been
“2.5G” capability. This is provided using technologies known
as general packet radio service (GPRS) and the enhanced data GSM
C HAPTER 2
The second major category of 2G cellular technologies
is the code-division multiple-access (CDMA) IS-95 family. These
systems were developed by the U.S. company Qualcomm and
are used primarily in the United States. They do not employ
the time-division multiple-access (TDMA) modulation approach
of 2G GSM. Instead, they carry multiple transmissions simultaneously by filling the channel with packets encoded for their
specific destination devices.
There is a family of upgrades for CDMA networks called
CDMA2000. This includes the CDMA2000 1x system, which
supports data rates up to 307 kbit/s. The CDMA Development
Group (CDG) reports their CDMA2000 subscribers as “3G”
users. Some operators and equipment manufacturers, however,
believe that the real equivalent to 3G is CDMA2000 1xEV (for
“evolution”), which is a higher-speed version of 1x. Within this
set of technologies are CDMA2000 1xEV-DO (data only) and
1xEV-DV (data/voice). Recent versions of EV-DO and EV-DV
support 3.1 Mbit/s downstream and 1.8 Mbit/s upstream theoretical data rates. Real-world rates are about half that speed.
The upgrade path from a 2G CDMA IS-95 network to a
3G CDMA2000 network is perhaps like the path from GSM
to GPRS or EDGE. Certainly, it can require similar attention
to handset, backhaul, and core network upgrades. But some
industry experts (and many CDMA advocates) have argued
that the base station upgrade requirements CDMA IS-95 to
CDMA2000 are easer and cheaper than those for GSM to
GPRS/EDGE.
It should be noted that, particularly in the case of GSM
systems, the 2.5G systems are merely way stations on the
migration path to full 3G systems (such as W-CDMA), which
are discussed in detail in the next chapter.
2.2.2.2
Broadband WLAN Technologies
To date, the most common broadband wireless WLAN
standard is Wi-Fi (IEEE 802.11). The 802.11 family currently
includes six over-the-air modulation techniques that all use the
23
same protocol. The most popular (and prolific) techniques are
those defined by the ‘a’, ‘b’, and ‘g’ variations to the original
standard. Wi-Fi uses different bands of radio spectrum. Perhaps
the most optimal use of Wi-Fi is to transform and extend the
capacity of standard wired Ethernet networks in public areas
like meeting rooms, training classrooms and large auditoriums.
In many places, of course, it has been adapted to provide localarea “hotspot” internet access. Wi-Fi generally provides a maximum of 54 Mbit/s data transfer to a maximum range of about
50 meters. The extension of Wi-Fi signal strength has emerged
as a niche development market in its own right, fostering the
growth of companies like AirMagnet (used in Antarctica to
extend Wi-Fi signals), among others.
In many North American and European countries, one
can find WLANs in public cafes, airports, university campuses, conference venues and other central, high-traffic locations. This is also slowly emerging in non-OECD countries.
For example, China’s largest shopping centre, the Super Brand
Mall in Shanghai, features the most extensive public WLAN
system in Shanghai. There are 67 access points, covering six of
the building’s 10 floors.
WLANs not only draw crowds, they also bolster the progressive, technological image that organizations, businesses and
institutes often seek to create. Moreover, WLANs contribute
to a sense of community (particularly in academic settings)
and provide new opportunities for generating access-related
revenues from people who are simply “passing through.” For
example, Switzerland offers a hot-spot service available in
every major city, allowing people in transit to benefit (at a small
fee, of course) from wireless broadband connectivity wherever
available network signals are detectable.
2.2.2.3
The Advent of WiMAX
Unlike Wi-Fi, which began life as a short-range networking technology, WiMAX is an infrastructure technology, on the
order of DSL or cable modem technology. It can provide the
same functionality as a router providing the backbone access to
a location. Individuals may connect to a WiMAX modem via a
wired Ethernet or Wi-Fi connection.
Based on a technique called orthogonal frequency division
multiplexing (OFDM), WiMAX is more bandwidth-efficient
than 3G technology or Wi-Fi. It supports high-throughput
broadband connections over long distances. WiMAX can be
used for a number of applications, including “last mile” broadband connections, hotspots, cellular backhaul, and high-speed
enterprise connectivity for businesses.
The prospect that mobile users may connect to WiMAX
“hot zones” directly is still in development, and mobile access
to WiMAX networks is not yet deployed. Meanwhile, genuine,
standards-compliant hardware has yet to appear, even for fixed
WiMAX installations.4 Figure 2.1 illustrates the growth forecast
for WiMAX equipment sales through 2009. There is currently
very little in the way of a track record by which to gauge the
cost-effectiveness of WiMAX as a technology.
Even the potential of WiMAX, however, raises interesting
regulatory questions. For example, how will the technology be
received in countries that have already auctioned off expensive
24
3G licences? Governments might not have as strong an incentive to promote WiMAX, particularly if it could enable competition with 3G services.5
Leaving aside the issues stemming from unlicensed usage
of spectrum for WiMAX (for example, potential interference), the growth of WiMax could depend on market forces.
It is not yet clear how much supply there will be from equipment manufacturers, nor is there a full picture of the level of
demand from consumers. If consumers want to use their PCs,
and these PCs are unable to leverage 3G spectrum to achieve
broadband connectivity in their neighbourhoods, WiMax
might be able to fill in the gap for ubiquitous wireless internet
access. On the other hand, the market model could simply be
scattered, shorter-range Wi-Fi hotspots, in tandem with mobile
data services.
Some transitional economies are moving faster than others
in providing wireless broadband connectivity. Considerable preliminary deployment activity is under way throughout Africa,
Asia and the Arab World. For instance, the Malian telephone
company Ikatel, a subsidiary of France Telecom, has contracted
a vendor to supply and deploy a WiMAX network in the capital city of Bamako.6 Meanwhile, in Saudi Arabia, broadband
wireless technology has been deployed to enhance data
transfer rates in the kingdom’s urban areas. Asian markets have been particularly active; representatives from
China, Japan and Korea (Rep.) have reached an agreement to
jointly develop future 3G technologies. Large-scale international manufacturers are also doing their part. Intel, for
instance, had announced its plans for WiMAX trials in
Malaysia, the Philippines and Thailand before the end
of 2005, and in Indonesia and Vietnam before the end
of 2006.7
2.2.3 Non-Terrestrial Broadband Platforms
Non-terrestrial systems (mostly satellite systems in the
current marketplace) are generally regarded as complementary to terrestrial broadband networks – and particularly useful
(or necessary) in remote areas where no terrestrial infrastructure exists. This view is reinforced by the high costs, limited
bandwidth and longer delays generally associated with satellite systems. Still, the number of subscribers for internet services provided by satellite operators is increasing. Aside from
TV distribution services and telephone connections, satellite
capacity is primarily used to connect internet service providers
(ISPs) to internet backbone facilities. It is less commonly used
for connecting individual users to an ISP, although this is certainly a growing market in rural and underserved areas.
Whether broadband internet access via satellite is the right
choice depends on the costs and performance of the alternatives.
The cost per megabyte of satellite access is decreasing, albeit
slowly. While the cost of operating the large, low earth orbit
(LEO) constellations limits their margins, regional geostationary systems should be able to reduce their costs significantly.
Prices on the order of USD 0.10 (10 cents) per megabyte have
been predicted. The open standard for digital video broadcasting, with a return channel via satellite (DVB-RCS) is increasing its market share and is expected to benefit both users and
C HAPTER 2
Box 2.1: Wi-Fi Beyond Hotspots
The explosive growth of Wi-Fi hotspots that provide wireless local connections in business complexes, homes, and public
spaces such as coffee shops and airport lounges has made Wi-Fi a household word. The IEEE 802.11 family of standards was designed by the data networking community for indoor, short-range, nomadic (can support walking speeds) uses. Wi-Fi has weak
handoff capability between access points. But the vision of Wi-Fi networks simply as a replacement technology to the physical
cables in the home or office LAN now seems constrained.
The amazing growth of Wi-Fi has driven economies of scale, such that the price for access points and end-user systems can be
as low as USD 50 or even less. And this attractive pricing, along with the spectrum licence exemptions that many countries offer
for Wi-Fi use, has led researchers to explore ways that Wi-Fi chipsets could be used in other networking environments.
A number of university research projects have been exploring modifications to 802.11 systems that would make them better
suited to long distance, point-to-point backhaul networking. This includes work at the Indian Institute of Technology campuses
in Kanpur and Chennai, as well as at the University of California, Berkeley in the United States. These projects have attempted
long-range (tens of kilometres) point-to-point backhaul hops using 802.11 systems.
The principle problem with the 802.11 standard, in this context, lies with the way multiple radios on the network contend
for transmission capacity. In technical jargon, the MAC layer uses a Carrier Sense Multiple Access (CSMA) protocol, which is
ill-suited for wide-area networks, although it is well-suited for local area networks. Research projects have developed new Wi-Fi
protocols better suited for long-distance, point-to-point networking. What may be emerging is a “rural network” extension to the
802.11 family that could directly compete with WiMAX.
More likely, however, is that Wi-Fi networks will complement WiMAX (and related networks). Wi-Fi hotspots can provide
nomadic broadband for the “last mile” (or “last metre”) distribution networks, while WiMAX or other technologies can take over
for backhaul and trunking.
Figure 2.1: WiMAX Growth, Forecast Sales, 2004-5
Worldwide Forecasted Sales of WiMAX Equipment
Source:
2005
$23 million
2006
$207 million
2007
$566 million
2008
$1.1 billion
2009
$1.5 billion
Sky Light Research, Sept. 29, 2005.
industry through lower costs, customer choice and equipment
interoperability. The market is, however, still dominated by a
few proprietary vendor solutions.
1
2
A satellite terminal consist of two parts:
An outdoor unit consisting of a transceiver and an antenna
that is placed in direct line of sight to the satellite, and
An indoor unit that provides the interface between the
transceiver and the end user’s communications system (a
computer or a local area network).
The aperture of the antenna decreases at higher frequencies, due to the reduction in parabolic antenna beam-width at
higher frequencies. Due to this, terminals operating at higher
frequencies are called Very Small Aperture Terminals or VSATs.
Most satellite systems for television broadcast and broadband data communications in operation today use portions
of the C-band and the Ku-band of spectrum. In the C-band,
C HAPTER 2
antennas are typically 2-4 metres in diameter, while in the
Ku-band they can be smaller than 1 metre, which also makes
it easier to direct the antenna. The lower-frequency parts of
the spectrum that the C-band and Ku-bands represent cannot
accommodate the data rates and traffic volumes demanded.
This has forced commercial satellite system operators to consider the Ka-band and V-band as well. But these higher frequencies present other challenges, including rain attenuation,
fading and signal scattering.
Independently of the system type, non-terrestrial wireless
broadband network topologies fall into the following categories:
–
Bent pipe star topology – This is characterized by a
large gateway earth station that transmits one or more
high-data-rate, forward-link broadcasts to a large number
of small user terminals. These broadcasts contain address
information that allows each user terminal to select those
25
transmissions intended for it. In the return direction,
the remote user terminals transmit in bursts at low-tomedium data rates to the gateway.
–
Bent pipe point-to-point – This topology calls for a
dedicated duplex connection, set up between a large gateway earth station and a single user terminal.
–
On-Board Processor (OBP) Switching – In this topology, the satellite rather than the gateway is the central node
in a star network. The satellite is connected to the gateway by one or more high-data-rate trunks. The on-board
processor de-multiplexes the uplink trunk into several
downlinks for different geographical areas, usually determined by the footprint pattern. The forward downlinks
contain messages for large numbers of user terminals, and
the destinations are identified by message headers. In the
return channel, the uplink transmissions from user terminals in one or more cells are multiplexed onto a downlink
trunk to the gateway.
The benefits of the “bent pipe” versus OBP are under discussion. In the bent-pipe system, the received signal is retransmitted without processing. While this scheme is less complex
and has performed well in the first phase of internet development via satellite, OBP promises better bandwidth efficiency
1
2
3
4
5
6
7
8
and true mesh connectivity. Several companies are conducting
OBP trials8.
2.3 Conclusion
Broadband networks are no longer a dream for the future.
Increasingly, they are at the centre of ICT development today.
The growth of converged, “triple play” offerings is coming to
dominate business and regulatory developments. This is particularly true in developed countries, but it is also a powerful
factor in developing ones. A full range of broadband technologies can and are being deployed in rural and underserved areas
of developing countries. Increasingly, regulatory conversations
around the world are laced with the fundamental question of
how can we ensure that all communities participate in, and
benefit from, the deployment of broadband capabilities?
The remainder of this edition of Trends seeks to answer
that question, and others that face regulators around the world
who are working to recalibrate their regulatory frameworks
to address broadband issues. The next chapter aims to help
regulators understand the technical dimensions of broadband
technologies and begin to discern how they will progress and
evolve, globally and in their own markets.
Chapter 3 will discuss the potential for stratospheric broadband platforms, which rely on sub-space aircraft such as airships to relay transmissions to and from the
ground.
The DSL family is often denoted by the abbreviation xDSL, in which the x denotes that there are various forms of the technology. Those forms are explored in more
detail in Chapter 3.
See: //http://www.cablemodem.com/specifications/re information about DOCSIS.
Gruman, Galen, Wireless Broadband’s long and winding road: Truly pervasive, high-speed mobile data services won’t happen overnight. Info World. Vol. 27, issue
39. September 26, 2005.
Gruman, Galen, “Wireless Broadband: The Long and Winding Road”, Infoworld, September 29, 2005. Link: http://www.infoworld.com/pdf/special_report/2005/
39SRwamobile.pdf.
“Ikatel to launch Wi-MAX”, Telegeography’s Commsupdate, November 29, 2005. Link: http://www.telegeography.com/cu/article.php?article_id=10078.
“Intel to Trial Wi-MAX in Indonesia, Malaysia, Philippines, Thailand, Vietnam”, PricewaterhouseCoopers Telecom Direct News, September 26, 2005.
These include Astrolink, SpaceWay and EuroSkyWay.
26
C HAPTER 2
3 UNDERSTANDING BROADBAND TECHNOLOGIES
Authors: Michael Best, Georgia Tech; and Bjorn Pehrson, Swedish Royal Institute of Technology
Each wave of technological development offers new
promise in the battle to bridge the Digital Divide. Most new
technologies are cheaper to deploy than legacy copper networks, and at the same time, they can deliver a full range of
ICT services, from voice to broadband applications and services. Many new technologies can also be deployed incrementally, even locally, rather than on the large scale of traditional
telecommunication networks.
Not only does this make deployment more affordable, it
opens the door to a whole new range of possible broadband
providers that can drive demand for broadband services. The
types of broadband providers enabled by new technological
developments include regional or private network operators,
small and micro entrepreneurs, as well as public institutions
such as universities, schools, libraries, post offices, local government offices, health facilities, and non-governmental organizations active in developing countries. This chapter identifies
and analyzes the promising new technologies that can help
promote broadband access in developing countries, particularly
in rural and underserved areas.
3.1 How Broadband Networks Are Designed
First, it is useful to introduce two concepts that help
explain different dimensions of networks: layered architecture and
network topology. These terms are important because they help
to understand the dynamics that determine incentives, viability and the potential for collaboration between entities that are
engaged in broadband deployment. In order to help regulators
make informed decisions about the array of technical options
before them, the following sections break out the various
aspects and layers of networks, providing definitions and technical specifications.
There are also useful comparisons between different types
of platforms, such as DSL, CATV, “broadband over power
line” (BPL) solutions, and fibre optical networks. These form
the basis of a “wire-line broadband roadmap,” outlined in Section 3.2, for regulators and policy-makers. That section also
provides detailed explanations of the various link layers (point
to point vs. point-to-multipoint access networks) and system
types. Attention is also given to wireless broadband systems, as
C HAPTER 3
well as options for non-terrestrial wireless broadband networks
when wire-line solutions are not feasible or applicable.
3.1.1 Layered Network Architecture
The concept of layered network architecture divides a network at any specific point into layers, each of which adds value
to the physical medium of communication. A layered architecture based on open standards is useful for several purposes:
–
Technical: to define physical and logical interfaces required
to connect different subsystems.
–
Commercial: to define the conditions under which a user or
a provider of value-added services can get access to services provided by a specific provider at a specific level.
–
Regulatory: to identify the value chains and define the roles
of actors providing services in different layers.
It may be necessary to regulate in order to promote competition and to allow service providers to buy services they
need in order to provide value added services at a higher level.
Rights-of-way, spectrum licensing, access to essential resources
and local loop unbundling, are all examples of areas in which
such regulation may be needed. The following subsections
explain the network layers in greater detail.
3.1.1.1
The Physical Layer
The physical layer identifies the transmission medium
– such as radio spectrum or copper or fibre wires – and specifies the mechanical and electrical interfaces that connect to the
medium for communication purposes. The most important
media include:
• Wireless spectrum: Existing spectrum regulations are often
too restrictive, from a public policy perspective, preventing new, innovative actors from entering the market. The
attempt to loosen these restrictions is clearly reflected
in the increasing reliance on licence-free spectrum in
the dynamic development of Wi-Fi (unlicensed usage is
explored more fully in Chapter 5 on broadband spectrum
management).
• Wire-line infrastructure: This includes fibre, copper cable
and coaxial cable. Wire and cable are essentially unlimited in supply and can be deployed and made available at
reasonable cost. In some countries, the provisioning of
27
passive infrastructure is an independent business, while
in others it is an integrated part of vertically integrated
operations. Wire-line regulation, too, can be more restrictive than necessary to promote broadband access. Access
to the physical medium is essential for entities that have
particular choices of transmission system or that want
to compete at the link level and above (for example, by
introducing a new competitive transmission technology).
Regulatory measures to promote access to both wire-line
and wireless broadband infrastructure are explored more
fully in the next chapter on regulators’ role in promoting
broadband.
3.1.1.2
The Link Layer
Access to link layer services is essential to service providers (or private network operators) that seek to build their own
networks without having to operate their own transmission
systems. With the physical layer as a foundation, the link layer
adds procedures for digital data transmission over the physical
medium, either point-to-point or point-to-multipoint. The link
layer includes access both to wireless spectrum and wire lines,
as well as to error control (such as automatic repeat requests
when check sums do not match), forward error correction
(based on the inclusion of redundant information coded in a
way that transmission errors can be corrected directly rather
than via retransmission requests), etc. Different link-level
technologies also have different properties that are important
for users, such as capacity, performance, security or privacy.
The properties of link-level technologies are different
and may be of different significance to operators, users and
regulators wanting to strike a balance between producer and
consumer interests. Links are implemented using transmission
equipment, and the most important ones to be discussed here
include:
• The most commonly used Wide Area Network (WAN)
technologies;
• Technologies for data access over legacy networks such as
digital subscriber lines (DSL) over the Public Switched
Telephone Network (PSTN);
• Cable modems over cable TV networks;
• Broadband over power line (BPL) transmission; and
• A range of both wired and wireless local and metropolitan area network (LAN/MAN) technologies, all of which
use the Ethernet frame data format according to the IEEE
802.3 standard1.
An Ethernet frame is the unit of data that is transmitted
between network points on an Ethernet network. Examples
of such wired networks include Ethernet with data rates from
10 Mbit/s to 100 gigabits per second (IEEE802.3). Examples of
wireless link level equipment include Wi-Fi (IEEE 802.11) and
WiMAX (802.16).
3.1.1.3
The Network Layer
The network layer provides mechanisms for addressing
and forwarding data. This chapter assumes that Internet Protocol (IP) is used for this purpose. The network layer is imple-
28
mented by network elements such as routers interfacing to
different link level technologies, link level switches and multiplexers, etc, to connect network hosts (servers and terminals).
This is the level at which all ISPs provide services.
3.1.1.4
The Transport Layer
The transport layer provides end-to-end connections
between user applications in network hosts. The central transport protocols include the connection-oriented Transport Control Protocol (TCP) and the User Datagram Protocol (UDP).
3.1.1.5
The Application Layer
In the application layer, the communication parts of a user
application, such as email or file transfer, Web access or database access, are implemented.
3.1.2 Network Topology
The network topology concept divides networks into
functional parts, including access, backbone and service networks and traffic exchange points where different service
providers exchange traffic. Each of these functional parts is
composed of the layers briefly described in Section 3.1.1. The
following subsections describe the functional parts of network
topologies.
3.1.2.1
Access Networks
Access networks are the links between end users and the
service providers’ networks, whether they are the first mile or
last mile (or last metre). First mile refers to a topology in which
the user or a local service provider – or perhaps even an apartment building company2 – owns the access network and connects to service providers using its own upstream links. A last
mile access topology denotes the operator’s ownership of the
access network.
Different link-level technologies used in access networks
have different properties, different sets of strengths and weaknesses and different value assessments by regulators, users,
operators and network owners. Depending on the geographical
context, the access network could be a LAN or a MAN. As will
be discussed in a later section, the technical solutions might
be different depending on who owns the access network. In
open regulatory environments, it could be owned by anyone
– a service provider, an organization, a municipality, or a user
agent or a neutral agent, such as a real estate owner or independent operator.
The physical layer infrastructure available for access networks includes the PSTN, cable TV networks, electrical power
networks, radio spectrum, and increasingly, fibre to the neighbourhood, office or home (this is often known as FTTP, for
“fibre to the premises”). On the user premises, the connections
to the residential or office gateway might employ dedicated
wiring, legacy copper telephone wiring, power-line or wireless facilities. The most common link level technology on user
premises is a local area network (IEEE 802), wired Ethernet or
wireless (Wi-Fi).
In an environment with more than one operator, local
loop unbundling is an important first step towards an open
C HAPTER 3
market. But this option is often frustrated in practical terms
by the fact that a dominant operator owns the infrastructure
and is being forced to lease it, essentially, to a competitor. A
more developed market structure may involve independent
ownership of the infrastructure, with mechanisms for service
providers to obtain direct access to users without intermediary
gatekeepers in the way.
3.1.2.2
Backbones
Backbones consist mainly of long-haul links that ISPs
can use to expand their service networks geographically, to
get transit to the Internet and to connect to regional traffic
exchange points. In some countries, an open backbone market
has emerged from the existence of parallel or complementary
fibre infrastructures deployed by different owners, including
telecom operators, power utility companies, railways, pipeline
companies and even municipal or regional governments.
3.1.2.3
Service Networks
A service network contains the ISP’s servers, which offer
Internet access to users via access networks, as well as transit
to the next tier of ISPs and peering with neighbouring ISPs in
the same tier.
3.1.2.4
Traffic Exchange Points
Traffic exchange points are used by operators to exchange
traffic through peering directly between service networks rather
than indirectly, via transit through their upstream providers.
An Internet exchange point (IXP) consists, in its simplest form,
of a link layer switch over which ISPs peer to exchange IP traffic. The exchange point improves network performance by
keeping local traffic local and minimizing transit costs for the
connected ISP. This is particularly important in areas where
the backbone consists mainly of satellite links with long delays,
high bandwidth prices and the remote ends on different continents.
This is a particular problem in Africa, which functions
as a large number of VSAT “islands” on the shores of other
Internet backbone “continents.” One solution to this isolation
is to interconnect IXPs, even via satellite links if better alternatives are lacking. This would cut the number of satellite hops
by making direct hops between IXPs rather than two transit
hops (sometimes more), plus perhaps a few transcontinental
and overseas passages depending on where the different transit
links happen to terminate. The main drawback of distributing
an IXP geographically by connecting local IXPs is the bundling
of the switching function and the long-haul link, which could
lead to unfair competition in situations where link capacity is
expensive.
In communities that only have VSATs available as gateways, it makes sense to have a local IXP keeping local traffic
local before sending it upstream via the satellite channel. This
is also valid for remote local communities that can take advantage of limited alternative fibre facilities – for example, links
used for metre monitoring in pipelines (water, oil, gas) or
power lines. IXPs can thus start appearing in the local access
networks, not just between service networks. The importance
C HAPTER 3
of IXPs is explored more fully in the joint ITU-IDRC Report
Via Africa: Creating local and regional IXPs to save money and bandwidth3.
3.2 A Wire-line Broadband Roadmap
To an increasing extent today, wherever possible, dedicated broadband networks are being deployed. Such networks
are typically based on Ethernet over fibre – in backbones, to
the curb, block or neighbourhood, and to the home or office.
They can interface to any other link level technology to take
advantage of existing infrastructure.
At the same time, new technologies have been developed
to provide broadband services over legacy networks, such as
DSL over the PSTN, hybrid fibre/coaxial (HFC) for cable TV
networks and broadband over power line (BPL).
The different link level technologies involved are discussed and analyzed in this section, in order to provide:
• A description of likely broadband upgrades based on existing infrastructures;
• A discussion of differences in quality of service and transmission rates among the various solutions;
• An indication of the kinds of infrastructure investments
that are required to deploy each technology (as a proxy for
cost); and
• A means of assessing the suitability of different network
options to various socio-economic and geographical contexts.
3.2.1 Upgrades to the PSTN
3.2.1.1
Dial-up and ISDN
Upgrading from the PSTN to ISDN requires a digital
network to the user premises. That, in turn, entails investment
in equipment, both at the central office and at the user end. If
already installed, ISDN is still an alternative for Internet access
in areas where more advanced services such as DSL, cable
modem or fibre networks cannot be used. But if ISDN is not
already in place, DSL appears to be a better investment because
it facilitates cheaper and higher-quality broadband service.
Unfortunately, ISDN is incompatible with some DSL systems
over the same infrastructure, so end users that have upgraded
from PSTN to ISDN actually may have to downgrade again,
before they can then upgrade to DSL.
Compared to dial-up service on the PSTN, ISDN is an
improvement – from both a bandwidth and reliability point of
view. Operators offer ISDN services including two channels
as a basic rate offering and 24 or 30 channels (depending on basic
PSTN type), as a primary rate offering. Available user equipment
consists of simple routers interfacing with the ISDN modem,
including 2 or 24/30 channels and an Ethernet interface for a
local area network on the customer premises. The equipment
supports automatic opening of new channels as needed when
the traffic increases, thereby providing from 64 kbit/s up to 1.5
to 2 Mbit/s connections. It is also possible to connect different
channels to different destinations.
29
On the operator side, the ISP typically leases a primary rate
connection (PRI) from the telecom operator. From the end user’s
point of view, the usage includes the call fee to the telecom
operator for each channel while connected, and the ISP fee,
which may be flat or connection-time based. Even in the best
of circumstances, with 56-64 kbit/s maximum bandwidth per
connection at voice tariffs, neither dial-up nor ISDN systems
are able to offer competitive broadband services.
3.2.1.2
Digital Subscriber Line (xDSL)
ADSL
Asymmetric Digital Subscriber Line (ADSL) is the most widespread DSL technology. The data channels use one frequency
band for a low-speed upstream channel (25 KHz to 138 KHz)
and another for a high-speed downstream channel (139 KHz
to 1.1 MHz). Data transmission speeds vary, based mainly
on the distance between the subscriber and the central office.
Some users cannot be reached at all by ADSL, because the distance to the central office is too great. In Denmark in 2004, for
example, about 5 per cent of households could not be reached
by any ADSL services, and only 70 per cent of the population
could access a 2 Mbit/s connection. More recent ADSL standards, such as ADSL2 and ADSL2+, promise improved capacity
and coverage.
VDSL
Very high-rate Digital Subscriber Line (VDSL)) is similar to
ADSL, except that it is optimized for shorter distances (3001 500 metres). Existing systems offer bandwidth capacities of
up to 52 Mbit/s by including more high-frequency bandwidth
in the copper cables and by deploying more efficient modulation. To extend its range, VDSL requires deployment of a fibre
optical backbone network to the curb, block or neighbourhood
(street cabinet). It also needs a power supply at the street cabinet, which is not required for service over the PSTN. This
increases deployment costs significantly. VDSL also has other
limitations, including interference from ADSL and AM radio
services. VDSL2, a standard under development, promises to
achieve bit rates of up to 100 Mbit/s.
Uni-DSL
One DSL for Universal Service (UDSL or UniDSL) is a new
variant of DSL, integrating all earlier DSL variants. It promises
aggregated bit rates of up to 200 Mbit/s, including 100 Mbit/s
symmetrical connections. While Uni-DSL gives operators the
flexibility to offer a range of connections, the higher data rates
cannot be offered on the existing PSTN infrastructure. UniDSL would require a fibre backbone infrastructure and would
use only the part of the existing subscriber line closest to the
user premises.
3.2.2 Upgrading CATV networks
A Hybrid Fibre/Coaxial (HFC) network combines a conventional coaxial cable TV network with fibre optic cables
installed between the head end and the curb, block or neighbourhood (interfaced by converters). An HFC network may
carry a variety of signal types, including analogue TV, digital
TV, telephone, and data. It increases the competitiveness of
30
cable operators in a manner similar to the way that Passive Optical Network (PON) upgrades reinforce the telephone industry
(PON is discussed in greater detail in section 3.2.4.2).
3.2.3 Broadband over Power Line (BPL)
Power line communication systems that use the existing
electrical power grid as a local loop for delivery of broadband
services are often referred to as Broadband over Power Line
(BPL). The typical power grid comprises generators, high-voltage lines (155-765 kilovolt or kV), substations, medium-voltage lines (1-40 kV), transformers and low-voltage lines (up to
400 V). High-voltage lines are unsuitable for BPL since there
are too many electromagnetic disturbances (noise).
Several organizations are working on standards to ensure
coexistence and interoperability between technologies, as well
as compliance with electro-magnetic compatibility (EMC).
The Institute of Electrical and Electronics Engineers, Inc.
(IEEE) has started work towards a “Standard for Broadband
over Power Line Hardware” (P1675),4 which is intended to
provide electric utilities with a comprehensive standard for
installing the required communication hardware on distribution lines. The standard is targeted for completion in mid-2006.
There are also working groups within the Special International
Committee for Radio-Electric Disturbances (CISPR),5 which
has produced relevant directives including EN55022 (European) and CISPR22 (international).
In Europe, standards include the low-power voltage 240volt usage and frequencies from 30 kHz to 150 kHz. In North
America, corresponding standards include the 120-volt grid
and set of frequencies above 150 kHz, as well. Power utility
companies often use frequencies below 490 kHz for their own
telemetry and equipment control purposes. BPL uses mediumvoltage power distribution lines (access BPL) and low-voltage
in-house wiring (in-house BPL).
Access BPL uses modems and couplers, which are inductive injectors wrapped around the power lines. Typically, a fibre
optic network connection from an ISP is terminated in an optoelectric converter and connected to a BPL modem at the utility substation, where the high-voltage lines are transformed to
medium-voltage distribution networks. The traffic is fed into,
and extracted from, the distribution lines via couplers.
The carrier supporting the communications signals can
share the same line with the electrical signals because they
operate at different frequencies. This is known as Frequency
Division Multiplexing (FDM) of telecom and electrical power,
with the BPL signal using frequencies between 2 MHz and
80 MHz.
Repeaters amplifying the signal and regenerating data must
be installed about every 300 metres between the power station
and the customer’s premises. Signals are terminated in a device
just before the transformation to low-voltage lines (110/220 V)
used inside the premises. BPL internal wiring facilitates home
networking by enabling devices plugged into wall outlets in
a building to communicate with each other over the existing
wiring. One formal industrial standard that serves as an industry reference point is HomePlug (www.homeplug.org), which
C HAPTER 3
offers specifications to operate in the frequency range of 4.521 MHz. Some systems in Europe operate at 10-30 MHz. Inhouse BPL and access BPL are not dependent on each other,
so either system can be employed with other technologies.
Examples of Access BPL system manufacturers include:
• Ilevo [www.ilevo.com] – offers products providing 200
Mbit/s and 45 Mbit/s bandwidth. Each power outlet is
an access point to the power line network. The Ilevo systems include a head end connecting the power grid to an
upstream ISP through any standard link-level technology,
different types of repeaters and a modem at the customer
premises. The frequency band used is 1-30 MHz.
• Amperion [www.amperion.com] – offers products that
deliver Internet connectivity via a wireless link called PowerWi-Fi (IEEE 802.11b) to an Ethernet port, instead of via
the in-house wiring. These systems operate over 3-35 kV
medium-volt lines, and they provide up to 24 Mbit/s of
throughput per injection point, depending on line quality
and equipment spacing.
3.2.4 Fibre Networks
So far, this chapter has discussed ways to extend existing networks or modify infrastructures to provide broadband
Internet access. This section takes another course: using fibre
technology to deploy broadband networks offering data as well
as voice and video services. In addition to a technical overview,
this section will discuss strategies for deployment of fibre in
developing countries – particularly in rural and underserved
areas, where penetration of wired telephone networks is low
and, in many cases, decreasing due to the popularity of mobile
telephony.
The wireless networks in many developing countries are
based on microwave backbones, which provide little support
for broadband applications. Many rural and underserved areas
lack any broadband communication infrastructure. Nevertheless, these regions do have users that demand broadband services. The lack of legacy infrastructure can clearly be turned
into a strength by leapfrogging over the need to build expensive, older-technology communication networks.
Instead, fibre deployment can be coordinated and even
shared with other infrastructure-dependent sectors, namely
power utilities (on which many ICTs rely to operate), railways, pipelines and roads. Many developing countries have
developed such strengths in recent years, after building political awareness through national ICT policies and infrastructure
plans. The most striking examples include Laos, Rwanda and
Tanzania. The availability of infrastructure creates new opportunities – but only if the regulatory environment allows entrepreneurs to take advantage of them.
3.2.4.1
The Physical Layer of a Fibre Network
An optical fibre is a hair-thin thread of glass that transports light waves with very low diminution over long distances.
Fibre is deployed in cables. Standard cables contain 24, 40 or
96 fibres. Cables can be deployed underground in conduits,
under water as submarine cables or strung between poles or
pylons. The cost of deploying fibre stems mainly from civil
C HAPTER 3
engineering work involved. The marginal cost of adding more
fibre cores in a cable is generally very low compared with costs
of other types of infrastructure.
Power utilities deploy fibre primarily for supervision, control
and data acquisition (SCADA) functions of managing the power
grid. But they are increasingly adding more fibre (at a very
low marginal cost) to lease to other parities. In new installations, utilities normally use a special ground wire with a fibre
cable in the core (an optical power ground wire or “OPGW”). Or,
where they already have power lines, they wrap fibre around
the transmission lines, in a process known as “SkyWrap.” Theoretically, then, every power grid substation – including those
in rural and underserved areas – can become a point of presence for access to fibre.
Signalling over optical fibre is accomplished by lasers as
transmitters and photo diodes as receivers. Standard data rates
are 1 gigabit per second (gbit/s), 2.5 gbit/s, 10 gbit/s or 40 gbit/s
in each stream. A 100 gbit/s prototype was presented at the 2005
European Conference on Optical Communication (ECOC).
Wavelength division multiplexing makes it possible to have up to
96 parallel data streams in a single fibre. The maximum total
capacity in a single fibre is currently in the 1-10 terabits per
second (tbit/s) range.
3.2.4.2
Optical Networks
The physical layer of optical fibre consists of two sublayers: (1) the passive fibre itself, without any signals; and (2)
an active optical network. In the passive fibre network, passive
optical splitters can be installed, creating a tree-structure infrastructure, similar to what is done with coaxial cable in cable
TV networks. This infrastructure is called a passive optical network (PON). It is used to implement point-to-multipoint links,
providing “broadcast” transmissions to end-users. The term
“passive” in this context means that the transmission of signals
from the central office to the customer premise equipment
does not require an external power source. Instead, PONs use
light waves for data transfer. The use of such trees to establish
links is discussed further in subsection 3.2.4.3. below.
Active optical networks are point-to-point infrastructures.
They include active (powered) optical components that provide routing, grooming and restoration of signals at the wavelength level, as well as wavelength-based services, by using
wavelength division multiplexing.6 This creates a purely optical
network infrastructure, before involving the electrical or digital
domains on the link level.
Not all fibre networks are designed using an optical network sub-layer in the physical layer. The main arguments for
introducing an optical sub-layer on top of the passive fibre,
before adding the digital communication link, are:
•
Better utilization of the installed optical fibre base, including the
possibility to resell capacity on a wavelength basis rather
than reselling an entire fibre. This argument mainly concerns the already-installed fibre base, since the marginal
31
•
•
cost of adding more fibre cores when deploying a new
cable is low.
Better network restoration capability after network failures,
since optical networks can perform protection switching
faster and more economically.
Reduced costs for the entire communication system, since the
distributed wavelength routing scheme decreases the cost
for cross-connects and only wavelengths that inject or tap
traffic at a node need an electrical network element at that
node.
3.2.4.3
The Link Layer of Fibre Networks
The dominant wire-line link level technology in fibre
access networks is Ethernet (IEEE 802.3). The 10/100 Mbit/s
Ethernet capability is standard in all new computers, including
laptops. A 1 gbit/s Ethernet capability is the standard interface
in most networking components used in access networks.
Regarding backbones, 10 gbit/s Ethernet has been in operation for quite some time in high-end network components. A
few 40 gbit/s backbones are in operation, and 100 gbit/s backbones soon will be deployed. In the backbone, the Synchronous Digital Hierarchy (SDH) and Synchronous Optical Network
(SONET)7 standards still dominate, because of the availability
of robust and reliable carrier-class equipment in those standards. The considerably cheaper and less complex Ethernet
technology is, however, making its way into the backbone,
reducing both capital and operating expenditures. Moreover,
Internet Exchange Points are now Ethernet-based. So the now
30-year-old Ethernet technology is finally expanding into all
parts of the network topology.
Point-to-point access networks establish independent
links between the user premises and service networks. They
offer maximum flexibility to all stakeholders, regardless of who
owns and operates the involved links. What can be termed the
operator-neutral model has been developed in Sweden. It is neutral in the sense that the passive access network infrastructure is
often owned by housing companies, condominiums or tenant
organizations and, in some cases, by municipalities.
In this model, access networks are connected to a shared
access network backbone. Any service provider can then connect its network gateway and offer services using the access
network. The access network backbone is designed so that
individual users in a housing area can select service providers independently of each other. Low-cost, standard Ethernet
multiplexers or switches are used to aggregate links from user
premises to the access network gateway, while preserving the
provider selection.
The operator-neutral model suits rural and underserved
areas well, since there are few traditional operators that are
likely to see a profitable business there. Given an adequate regulatory environment and appropriate technologies, local entrepreneurs that know the local market opportunities can provide
local services and connect users to the network points of presence (POPs) of the larger service providers.
Turning to point-to-multipoint networks, the IEEE
802.3ah Ethernet in the First Mile standard and the ITU-T G.984
32
GPON standard have made Ethernet the preferred protocol
also among the traditional, vertically integrated telecom operators. As discussed in Section 3.2.4.2, a passive optical network,
or PON, is a point-to-multipoint technology similar to cable
networks, but fibre-based. By introducing passive optical splitters and couplers, a tree structure is created from an optical
line terminator at the central office to optical network terminals
(ONTs) at a number of customer premises. Downstream data
is broadcast to the terminals, each of which looks for a matching address at the protocol transmission unit header. Upstream
traffic is coordinated using a TDMA protocol, in which dedicated transmission “time slots” are granted to each terminal.
The main fibre can operate at 155 Mbit/s, 622 Mbit/s
(Broadband PON or BPON managing up to 16 ONTs) and
1.25 gbit/s or 2.5 gbit/s (Gigabit PON or GPON managing up
to 32 ONTs). Bandwidth allocated to each customer from this
aggregate bandwidth can be static or dynamically assigned in
order to support voice, data and video applications. The terminal can provide all the appropriate interfaces. A single fibre,
meanwhile, can serve 16, 32, or more buildings through the
use of passive devices to split the optical signal and PON protocols to control the transmission of signals across the shared
access facility.
From an economic point of view, PON saves on the cost
of fibre and equipment at the central office/head end, compared
to using point-to-point connections. Any savings should, however, be weighed against the weaknesses of point-to-multipoint
technologies, which are similar to those of broadband cable
networks. For a regulator pushing an open regulatory regime,
the lock-in effect of PON is less attractive, since the topology
of the physical medium makes it impossible to separate users
and thus prevents local loop unbundling on the physical and
link levels. There is no unique path between the central office
and a single user, because passive splitters are used to build the
physical infrastructure, as in a cable network. Active optical
networks, on the other hand, use switches, so end users can
be separated.
For the operator, meanwhile, a PON is more complex to
plan, manage and upgrade than point-to-point links. For the
user, the performance depends on traffic from other users
(because of bandwidth sharing) unless traffic control of individual connections is introduced. The broadcasting technique
also raises concerns about security and privacy. Policy-makers
and regulators seeking to promote new fibre backbone deployment should weigh carefully the costs and benefits of these two
fibre network options in order to best meet the ICT development goals of their country.
3.2.4.4
Fibre Deployment in Developing Countries
Fibre deployment is taking off in developing countries.
The technologies exist and are not that expensive. In many of
the developing countries deploying fibre, universities are at the
forefront, establishing national research and education networks (NRENs). Bangladesh, India and Pakistan, for example,
all have national fibre backbones. NREN examples include:
C HAPTER 3
Table 3.1: Relative Functionalities of Broadband Wireless Access Technologies
Technology
Used frequencies
Supported data rates
Cell radius
Notes
EDGE
850 / 900 / 1 800 /
1 900 MHz
Up to 384 kbit/s
1 km
Deployed globally; extensive availability of
terminals
WCDMA
1 900 / 2 100 MHz
Up to 2 Mbit/s
0.4-2 km
Large-scale deployment; widely-supported
by vendors
HSDPA
1 900 / 2 100 MHz
Up to 14 Mbit/s
2 km
1xRTT
450 / 850 / 950 / 1 800 /
1 900 / 2 500 MHz
Up to 144 kbit/s
Up to app. 50 km
2.3 GHz
Up to 2.4 Mbit/s
Up to 15 km
1.9 / 2 / 3.4-3.5 GHz
Up to 7 Mbit/s
29 km
Deployed in a few countries, e.g. New
Zealand, Australia, Portugal, etc.
WiFi
2.4 GHz
Up to 11 Mbit/s
100 m
Widely deployed globally; backed by major
vendors with range of terminals available
WiMAX
3.5 GHz
Up to 75 Mbit/s
Up to 50 km with
line of sight
Believed to be optimal solution for fixed
wireless access and for pushing broadband
into rural areas
1xEV-DO
UMTSTDD
•
•
•
•
•
•
•
The Pakistan Education and Research Network (PERN)
connects all public universities in Pakistan.
Laos has a national fibre infrastructure reaching all provincial and district capitals. There is an Internet exchange
point (IXP), to which all ISPs, except the incumbent, are
connected. The universities are in the process of setting
up an NREN, and they participate in an European Union
funded regional academic backbone programme, dubbed
TEIN2, that connects NRENs in the ASEAN countries.
For a reported cost of USD 50 million, a Kenyan network
provider has commissioned the installation of a 1,140 km
optical fibre network by the end of 2006. The optical fibre
technology will give mobile operators, ISPs and fixed-line
operators a core network extending from the coastal city
of Mombasa in the southeast to the country’s western
border.
In Rwanda, Terracom is in the process of deploying a fibre
network to all schools and other priority groups. Rwanda
has connected all of its ISPs at an (IXP). Meanwhile, academic institutions are in the process of organizing an
NREN.
Both Tanzania and Mozambique have a mix of multiple
fibre owners. Both have IXPs and are in the process of
establishing NRENs.
Malawi and Zambia rely on power utilities for fibre
deployment. Both are in the process of deploying IXPs,
and the universities are in the process of setting up
NRENs in both countries.
Bolivia has a national fibre backbone, an IXP and an
expanding NREN.
C HAPTER 3
Enhancement to WCDMA; limited
deployment in Japan
Widely used as fixed wireless solution
Deployment mostly concentrated in North
Asia; widely supported by vendors
3.3 Broadband Wireless Access (BWA) Networks
Wire-line broadband deployments, of course, are not the
only options. A range of terrestrial wireless solutions are on
offer or on the horizon. The development of BWA solutions
has been marked both by significant technological progress and
substantial “weeding out” of some market players. BWA solutions can be divided into four primary sub-categories, two of
which are also often labelled as 3G or IMT-2000 technologies
(W-CDMA and CDMA2000). The other two major families,
WiMAX and 802.20, are growing in use and maturity in parallel
with the development of their standards. While the 3G systems
have evolved out of the mobile telephony sector, the other two
technologies have emerged from the data networking sectors
and often offer fixed wireless solutions (although they, too, are
moving towards mobility).
This section explores the migration paths of the four
families of broadband wireless access technologies, picking
up from the 2G and 2.5 solutions explored in Chapter 2. The
technologies in this chapter represent the immediate future of
broadband development.
3.3.1 The GSM Migration Path
The 3G migration path from GSM, known as wideband
CDMA (W-CMDA), is an incremental upgrade from GSM’s
2.5G networks. W-CDMA networks, sometimes called UMTS,
have been initially deployed in Europe and Japan. Standard WCDMA systems can support up to 2 Mbit/s, while an enhanced
version, called HSDPA, allows downlink rates of up to 14
Mbit/s by using higher modulation rates and other advanced
techniques. With 2.5G base stations and available backhaul
networks in place – and with spectrum secured – the upgrade
33
Box 3.1: A Radio Transmission Primer
All wireless networks communicate via electromagnetic energy, which is described by wavelength or frequency. Since electromagnetic energy traces a sine wave, the frequency specifies the number of humps (or troughs) per second of time and this unit
is called the Hertz (or just Hz). Radio waves between 30 MHz (million Hz or megahertz) and 20 GHz (thousand million Hz or
gigahertz) are usually used for data. Lower frequencies are used mostly for broadcast services such as FM and AM radio (though
in some cases these are also used for data). Energy much above 20 GHz is not very suitable for data over long distances, as it is
easily absorbed by water vapour in the atmosphere.
The frequency of electromagnetic energy tells a lot about its usefulness for communications. Relatively low frequency energy
– AM radio broadcasting frequencies, for example – travels as ground waves, literally hugging the earth’s surface. This allows the
radio waves to travel past the horizon dramatically, extending their distance. These waves are not easily absorbed or reflected by
objects such as trees, buildings, and the like. Relatively low frequency energy also can be bounced off of the ionosphere, dramatically extending the distance travelled (see below).
By contrast, higher frequencies, in the spectrum known as “microwave,” generally require line-of-sight (LOS) transmission
and cannot travel much past the horizon. The LOS requirement can be mitigated, as seen in some emerging 802.16 technologies,
through sophisticated multi-path approaches, in which the signal reaches the receiver not necessarily in a straight line from the
transmitter but instead reflecting off one or more objects, creating multiple paths from the sender to the receiver.
Given these transmission or “propagation” properties, it would seem that the lower the frequency, the easier and better the
transmission. But in thinking about broadband digital data transmission, there is one major advantage of higher frequencies: the
higher the frequency, the more data that can be transmitted. In fact, typically 1-4 bit/s of data can be encoded for each cycle-persecond of the radio wave. Note that this not the same as the data rate delivered to a user, because many of these bits are put to
other purposes, such as signalling, error detection and correction. Broadband wireless networks are all situated in the microwave
spectrum in order to take advantage of much higher signalling rates. But they all must do battle with the relative difficulties in
signal loss and fading.
Figure 3.1: Comparative Prices for Mobile Data Services
Price per Mbit/s, for selected operators in selected countries, June 2004, in USD, at different monthly usage thresholds, compared with NTT DoCoMo’ s unlimited
usage monthly price
Operator (country)
1 Mbit/s
10 Mbit/s
100 Mbit/s
1 Gbit/s
GPRS
1.46
1.22
1.22
1.22
T-Mobile (Germany)
GPRS
10.68
2.62
2.09
0.23
TIM (Italy)
GPRS
1.82
1.82
1.82
1.82
Telefonica (Spain)
GPRS
7.29
7.29
1.46
1.46
Vodafone (UK)
GPRS
13.71
3.86
2.15
2.15
NTT Do CoMo (Japan) (package)
W-CDMA
9.14
3.92
1.35
0.07
NTT Do CoMo (flat-rate)
W-CDMA
35.00
3.50
0.35
0.04
Orange (France)
Source:
34
Technology
3G Mobile, ITU research
C HAPTER 3
Box 3.2: TD-SCDMA – A Chinese Standard
Time-Division Synchronous Code-Division Multiple-Access (TD-SCDMA) is a 3G standard created by the Chinese Academy of
Telecommunications Technology, working with equipment vendors such as Siemens. Similar to the WiBRO initiative described
below, TD-SCDMA offers an interesting illustration of a way that the public sector can collaborate with the private sector to create
a local network standard. In fact, the Chinese have explicitly positioned TD-SCDMA as a way for the country to avoid dependence
on “Western technologies.” And with China having more mobile subscribers than any other nation in the world, the country has
a sufficient market size to support its own standard.
The TD-SCDMA standard is an evolution from the GSM standard in much the same way W-CDMA is. Field tests have
shown the system to work at vehicular speeds and at a 21 Km distance from the base station. Data rates are published as ranging
from 1.2 kbit/s up to 2 Mbit/s.
The fact that the published data rates are so broad is more than a curiosity, as technical trials have shown a disappointing
delivered capacity. The government has called for an “intensive industry-wide effort” to deliver on the technology and to make
it competitive.
Figure 3.2: Migration Paths from 2G to 3G
Figure 2 — From 2G to 3G
Evolution of mobile systems to 3G
TDMA
GSM
EDGE
GPRS
WCDMA
PDC
cdmaOne
CDMA2000 1X
2G
First step into 3G
3G Phase 1
Evolved 3G
Voice and up to 14.4 kbit/s data
64 – 144 kbit/s
384 kbit/s – 2 Mbit/s
384 kbit/s – 10 Mbit/s
Actual user data rate 14 kbit/s
20 – 60 kbit/s
60 – 120 kbit/s
Over 384 kbit/s
Time
2000/2001
from 2.5 to 3G systems can be done incrementally and with
relative ease. For instance, 3G services can be offered on a cellby-cell basis and only in those areas with sufficient subscriber
demand. In practice, however, 3G upgrades are done across an
entire region in order to provide consistency in service levels
for mobile customers and to allow broad marketing and sales
plans.
3.3.2 The CDMA Migration Path
There is a family of 3G upgrades for CDMA networks
that are called CDMA2000. This includes the CDMA2000
1x systems which support data rates up to 307 kbit/s. The
CDMA Development Group (CDG) refers to all CDMA2000
subscriber numbers as “3G” users, but the real equivalent to
3G is in fact CDMA2000 1xEV (for Evolution), which is a
C HAPTER 3
CDMA2000 1XEV
2001/2002
2003
higher-speed version of 1x. Within this set of technologies are
CDMA2000 1xEV-DO (data only) and 1xEV-DV (data/voice).
Recent versions of EV-DO and EV-DV support theoretical
data rates of 3.1 Mbit/s downstream and 1.8 Mbit/s upstream.
Real-world rates are about half that speed.
The performance and market adoption of 3G networks
(CDMA2000 and W-CDMA) has, to date, been disappointing.
This is partly because of high prices associated with network
upgrades and the extraordinary amounts paid for 3G spectrum
in some areas (mostly Europe). There also is a relative lack of
suitable applications to motivate subscriber adoption and handset upgrades. But the 3G vision remains compelling: ubiquitous Internet access enjoyed while moving at vehicular speeds,
with broadband bandwidths and latencies that are comparable
to DSL or cable modem services.
35
Box 3.3: 802.16 Extensions in the Works
802.16a
Works in 2-11 GHz range and supports mesh deployments.
802.16b
Increases the amount of spectrum that can be used in 5-6 GHz range, and provides Quality of Service guarantees.
802.16c
Works in a higher frequency range of 10 to 66 GHz.
802.16d
Brings improvements to 802.16a; intended to supplant 802.16 and 802.16a.
802.16e
Supports mobile devices.
Box 3.4: The Evolution of a “Southern” Solution
Broadband corDECT is an important, although incremental, improvement on the corDECT system designed at the Indian
Institute of Technology, Madras. The initial DECT system, based on the European standard originally developed for cordless
telephones, provided LOS connectivity at a maximum of 70 kbit/s. One significant strength of this system has been its relative
low cost.
While the benefits of non-line-of-sight (NLOS) systems are palpable, there is no immediately available technique to develop
an NLOS DECT-based system. This is because of the power restrictions on the spectrum used by DECT radios. Broadband corDECT is able to turn this “bug” into a feature by taking advantage of the line-of-sight requirement to enhance capacity. Through
spatial reuse of spectrum, enhanced modulation levels, and the use of radio wave polarization techniques, Broadband corDECT
delivers 256 kbit/s – and, ultimately, as much as 512 kbit/s – of dedicated bandwidth to each user. Costs are considered affordable
at around USD 150 per subscriber, including base station and customer equipment costs.
Source:
Midas Communications
3.3.3 The Evolution of WiMAX
A range of technologies fall under the WiMAX moniker,
including those that conform to the emerging set of IEEE
802.16 standards. WiMAX systems promise to:
–
Have very high capacities (up to 134.4 Mbit/s in a 28 MHz
channel),
–
Travel long distances (50 km or more),
–
Not require line-of-sight,
–
Work at vehicular speeds (under the 802.16e extension),
–
Enjoy high spectral efficiency (under the 802.16a/d extension), and
–
Be inexpensive (with base stations in the $10,000 range).
It sounds like a broadband wireless “dream come true,”
except that not all of these extensions have worked yet in the
real world. Moreover, not all of these desirable qualities can be
enjoyed at the same time with the same network (for example,
as of yet there is no cheap, efficient, high-capacity system that
works at vehicular speeds). Still, WiMAX systems show great
promise for the provision of broadband Internet access services, especially in remote areas or wherever fully ubiquitous
access, at vehicular speeds with seamless handoff, is not a high
priority.
The 802.16 standard and its extensions were still being
finalized at the time of publication, but a number of vendors
already had begun offering broadband metropolitan area network (MAN) technologies that will comport with the 802.16
standard as soon as it is stable. One example is the Canopy
system from Motorola. Canopy transmissions can travel 50
km in a single hop, providing 10 Mbit/s of shared bandwidth.
36
Outdoor access points can list for USD 1,000, with customer
premises equipment costing USD 500.
Samsung and LG Electronics of Korea (Rep.) have developed a WiMAX-styled technology called WiBro (for “wireless
broadband”), which is designed for the 2.3 GHz band. It offers
512-1,024 kbit/s per user, and allows users to travel at nearvehicular speeds (around 60 km per hour). The system has
emerged with assistance from the government of Korea (Rep.),
which was eager to see a locally produced technology and had
promoted WiBro as the basis for the 802.16e mobile WiMAX
standard. Other important stakeholders, however, were not
supportive of WiBro as a standard-setting technology for technical reasons. The 802.16e standardization process could have
been weakened if two competing technologies were to emerge.
But at this point, companies have agreed to converge on a
shared single standard.
The European Telecommunications Standards Institute
(ETSI) has also developed broadband metropolitan area network standards under the name HiperMAN. Like WiBro and
other related technologies, these systems allow for long-distance transmissions (over tens of kilometres) and high bandwidth (up to 280 Mbit/s at each base station). The WiMAX
Forum has been working with the HiperMAN, WiBro, and
IEEE 802.16 standards to try to ensure interoperability among
all of these various systems.
3.3.4 The New Kid on the Block: 802.20
WiMAX started as a fixed wireless technology and has
since been evolving to support mobility under the .16e extension. The IEEE’s 802.20 standard, however, originated explic-
C HAPTER 3
Box 3.5: Mesh Networks
A mesh network interconnects communication nodes to pass messages directly between each node, without a dedicated
intermediary (i.e., a server). The communication nodes used in a mesh network can be of any type – from handheld sensors to
desktop computers to Wi-Fi routers.
The difference between mesh networking between clients (also known as multipoint-to-multipoint or peer-to-peer networking) and traditional networking is that in the latter, a dedicated router is required to pass all messages from one client to another.
Mesh networks still frequently make use of dedicated routers, but they also can be configured without them. The arrangement
(topology) of client communication devices (nodes) is also very flexible. The key is that each node is capable of talking directly to
its neighbours (within range), whether that node is a client (computers, sensors, kiosks) or infrastructure (routers, access points,
gateways).
There are several reasons to choose a mesh network over a traditional one, including greater scalability, extensibility, resiliency,
and physical infrastructure advantages. Scalability refers to the flexibility of a mesh network to choose the path of a message being
routed. When an additional node is added to a traditional network, it has to be able to communicate with a dedicated router, which
either increases the load on an existing router or requires installation of a new one. On a mesh network, however, the additional
node not only consumes routing capacity but also adds its own capacity to the network. This means that the routing capacity and
aggregate throughput within the network grows as the network grows, rather than starting at some limit and decreasing, which
is what happens in a traditional network.
Extensibility is a very important advantage of mesh networks – particularly wireless ones. Extensibility stems from the fact
that any node can have its message hop from one point to another via a peer node. In a traditional (point-to-multipoint) wireless
network, being out of range of a dedicated access point means a loss of connectivity. But in a mesh network, connectivity can be
maintained as long as there is another client node within range that can be used as a bridge to eventually reach the access point
or destination. There is no fundamental limit to the number of hops a message can make before reaching its destination, but implementation details mean that, as the network extends farther and farther outward, it will eventually be beneficial to add a new
dedicated access point.
Another advantage of mesh networks is their tolerance of failures. Typically, in a mesh network there will be more than one path
available for data to take. A traditional network only offers a single path. If, in a traditional network, a router suffers a failure, the
nodes connected to it lose all connectivity. Another benefit of the multi-path character of mesh networks is that the “network
diameter” – the minimum number of hops separating any two communicating nodes – can be smaller than in a traditional network.
This can result in reduced latency on the network.
There are drawbacks to mesh networking. First and foremost is that each communicating node needs to be willing and able
to route traffic. This activity requires computational and electrical power and can slow down the appliance or unduly drain its
battery. In addition, mesh networks require additional preparation and setup of each client and may be harder to maintain than
a more centralized, traditional network.
One market in which mesh networking could make significant inroads is rural access. A mesh network provider based in a
metropolitan area could offer services in remote areas by “piggybacking” connectivity over a series of subscribers in the direction of
the end-user. Data traffic on the outer edge of the network – in a remote village, perhaps – would only need a wireless connection
strong enough to reach the next, closer subscriber to the metropolitan area. This second subscriber would then pass the traffic to
another, closer subscriber and the process would continue until the traffic reached the backbone Internet connection. By using all
subscribers as transit points, the mesh network can quickly reach distant areas with relative ease.1
1
ITU Internet Reports: The Portable Internet, 2004, see http://www.itu.int/publications/bookshop
itly as a mobile broadband technology. The fact that its original
design requirements included mobility should prove beneficial to the standard compared with others, which have had to
add mobility on as an extension. Even so, the 802.16 family
of WiMAX systems has a first-mover advantage and, at present,
stronger industry support from major players.
Through its recent acquisition of Flarion Technologies
and their Flash-OFDM technology, Qualcomm has just demonstrated its support of the emerging 802.20 standard, which
has not yet been finalized.
C HAPTER 3
3.3.5 OFDM Technology
Orthogonal frequency-division multiplexing (OFDM) is emerging as a leading technology for very-high-bandwidth wireless
connectivity. As the speed of wireless services increases, it
increases the need for more and more radio spectrum, which
is expensive and hard to acquire. In turn, this increases the
importance of spectral efficiency, which refers to the number of
bits that can be encoded into a single radio cycle. OFDM-based
technologies, including WiMAX, enjoy spectral efficiencies of
around 4 bit/s per hertz. This compares favourably with the
37
Figure 3.3: A mesh network topology versus a traditional network
802.11d standard, for instance, which has a spectral efficiency
of less than 2 bit/s per hertz.
OFDM works by segmenting available spectrum by frequency and then carrying a portion of the user’s data on each
of these frequencies. Each frequency is unique and does not
overlap, making it orthogonal to the next. This ensures that
there is no interference between the various tones. This technique, along with other sophisticated improvements in digital
signal processing, has produced an efficient and speedy network technology.
Flash-OFDM can deliver capacity of about 1 Mbit/s downstream and 500 kbit/s upstream while stationary. Available
bandwidth is diminished, however, when moving at vehicular
speeds. One significant strength of the Flash-OFDM system is
its spectral efficiency, which is about 4 bit/s per hertz (a similar
efficiency is planned for the 802.16a OFDM extension). This
is a great advantage when spectrum is scarce or expensive. In
some rural and underserved areas, where microwave radio
bands are relatively underutilized, this advantage is not as compelling.
Another early system intended to track the emerging
802.20 standard is iBurst, which is based on radio technologies
(and in particular smart antenna designs) produced by U.S.based ArrayComm. Early deployments of this mobile broadband system have gone up in Australia and South Africa.8
3.3.6 Integrating Heterogeneous BWA Networks
What this BWA roadmap makes clear is that there are a
number of complementary and often-competing standards,
standard-setting institutions, proprietary offerings and vendors.
Considerable effort is being put to interoperability and merging
these various systems. But it still is likely that many enterprises
will end up with a heterogeneous collection of wireless networks. Moreover, the functionality of many of these systems is
converging (See Box 3.6).
38
Ongoing research is focusing on ways to integrate the various network technologies. For instance, the Third Generation
Partnership Project (3GPP) has been studying systems to interconnect 3G systems with WiMAX or Wi-Fi networks. Issues
have included hand-off as well as authentication, authorization and accounting (AAA) and other considerations. As the
capabilities and purposes of fixed wireless and 3G wireless
networks converge and approach the performance of wire-line
networks, perhaps it will become easier to make a strategic
choice between these families of technologies.
3.3.7 Civil Engineering Costs for Broadband
Wireless
As long as Moore’s law holds true, there will be a constant, exponential growth in the performance and affordability
of computer technologies. Regrettably, however, Moore’s law
has yet to apply to the performance or cost of steel, cement,
or labour. As a result, the “civil engineering” costs for outdoor
radio networks – the costs of radio towers, cement foundations,
and related requirements – are beginning to outdistance the
cost of the solid state radio equipment.
Consider the cost of radio masts or towers, which are
essential for fixed wireless and 3G radio networks. Without
towers, it is impossible to position signals above foliage and
buildings that might obstruct them, to extend the horizon, or
to meet other transmission requirements. In most major cities,
the skyline is littered with these towers. They range from
masts on the order of a metre in height, positioned on the roof
of a customer’s premises, to 10-metre cellular network towers
and 100-metre (or higher) towers populated with high capacity
point-to-point microwave antennas and, perhaps, broadcast TV
or radio facilities.
The cost and effort required to put up a radio tower can be
influenced by several factors, including local climate (whether
there are high winds or frequent icing) and the intended load-
C HAPTER 3
Box 3.6: Convergence of WLAN and 3G Networks
There is today no single, optimal broadband technology. Each major family of technologies features strengths and weaknesses.
Figure 3.4 provides a preliminary decision matrix that can help develop strategic choices among the technology types. Notwithstanding these technology differences, there is significant convergence among the various offerings towards a unified set of design
characteristics for an ideal broadband wireless network. These characteristics would include:
– High bit-rates in an all-IP environment, including IP Version 6 (IPv6) support
– End-to-end QoS
– Multimedia support
– Mobility at automobile and train speeds
– Seamless session management
– Security, security, security
– Support for flexible and dynamic spectrum and interference management (including software-defined radios)
– Advanced authentication, authorization, and accounting protocols.
These goals are more than a pipe-dream. The fixed wireless and 3G families of networks are converging on each other, and
beginning to converge on this network design “wish list.”
Figure 3.4: The Convergence of WLAN and 3G Technologies – A Matrix
Current WLAN Strengths
Emerging WLAN Strengths
• Relatively inexpensive
• Mobility
(e.g. mobile WiMAX 802.16e)
• Data ready and high-bit rates
• QoS (e.g. WiFi 802.11e)
• VoIP
Emerging 3G Strengths
Current 2G strengths
• Ubiquitous
• Multimedia support
• Mobility
• High data rates
(e.g. W-CDMA)
• QoS guarantees
ing of the tower (for example, one small antenna or many large
ones). Towers are designed with these factors in mind, and a
choice must be made between three principal support strategies:
•
Free-standing,
•
Bracketed against an existing structure, or
•
Supported with guy wires.
By far the most important factor in determining the cost
and complexity of a tower design is its height. The height
requirement largely dictates the type of tower that is needed,
the foundation strength and the requirements for guy wires,
among other things. Grounding is important, particularly
in areas that experience seasonal electric storms. Lightening
strikes can spell the death of antennas and radio electronics.
In order to protect the equipment, lightening attractors can
be affixed at the top of the tower. The principle of a lightening attractor is simple: it should be higher than the equipment
being protected and it should offer less resistance to the flow
C HAPTER 3
of electricity than “competing” local options (such as the radio
equipment). Fixing an attractor above the radio and antenna
is easy enough. But ensuring low electrical resistance requires
careful attention to grounding. Especially in rural and underserved areas – where, for instance, a relay base station may be
positioned – industrial grounds are not available and it is necessary to dig an earth pit to ground the arrestor. This activity
can add substantial expenses to the costs.
Figure 3.5 shows empirical data for the tower and labour
costs in the United States and Ghana, as well as grounding
costs in the United States. A simple regression performed on
these data points yields the linear cost models that are depicted
visually below. Results indicate that:
•
In Ghana a 3-metre tower with installation and grounding
costs just under USD 220, and an additional USD 270 for
each additional metre.
39
Figure 3.5: Empirical Data for Tower, Installation, and Grounding Costs
Figure 3.6: Linear Model of Tower, Installation, and Grounding costs (USA/Ghana)
40
C HAPTER 3
Figure 3.7: Radio Tower Types and Heights
A simple radio pole, perhaps just a few meters high (left); a free-standing radio tower used for a cellular phone network, perhaps 10 meters high (centre); a guyed
radio tower holding various antenna, perhaps 100 meters high (right).
•
In the United States a 3-metre tower with installation and
grounding costs nearly USD 1,360, and almost an additional USD 770 for each additional metre.
Of course, these models are just estimates to help to clarify and guide thinking. As a point of comparison, for instance,
Motorola Canopy outdoor access points can list for approximately USD 1,000, while their subscriber modules can be
under USD 500. Clearly any cost analysis of wireless broadband equipment should consider such civil engineering costs,
since these can dominate costs for radio equipment.
3.3.8 Stratospheric platforms
Stratospheric platforms operate within the stratosphere
portion of the atmosphere, under aerodynamic conditions.
High- and low-altitude platform stations (HAPS or LAPS) offer
the coverage benefits of satellites at costs close to fixed infrastructure.9 A HAPS platform may be an airship soaring in the
stratosphere at an altitude about 20 km – well above normal
aircraft but below orbiting satellites. There are no commercial
HAPS services yet, but active research and development efforts
are under way. LAPS services may employ an airship kept stationary at an altitude of about 3 km. Commercial LAPS services are available, especially targeting temporary needs such as
emergency response situations or sports events.
The key markets for LAPS and HAPS will likely be rural
and developing areas that are underserved by traditional infrastructure. But they could also play a key role in conjunction
with newer wireless technologies such as WiMAX. Since LAPS
and HAPS are potential platforms for delivering connectivity
for a range of wireless systems, their radio equipment could
make use of the most current technologies to provide fast line-
C HAPTER 3
of-sight connectivity. Since the line-of-sight requirement could
be met for many applications, the frequencies and corresponding transmission speeds could be much higher.10
3.4 A Decision Framework for Broadband
Facilities-based competition is a desirable element for any
nation’s broadband market, but it simply may not exist in all
(or any) parts of all countries. For instance, some high-income
metropolitan areas may have multiple wire-line broadband providers (a cable TV system operator and a DSL operator) along
with multiple mobile operators. Some areas may be served by
smaller, local entrepreneurs or even municipalities offering
WLAN hotspots. Each of these would compete to provide the
most cutting-edge data solutions. But in some low-income
areas – and especially in rural and sparsely populated areas
– this level of competition may not be commercially viable.
Initially, policy-makers and regulators may succeed
in inducing only one broadband provider to enter rural or
sparsely populated markets, while leaving the door open for
other entrants once demand for broadband services has been
established. In these environments, therefore, strategic choices
may have to be made between various network families. In
other words, some contexts may commercially support facilities-based competition (or, said another way, “complementary”
networks), while other regions may face a choice of one network over the others.
As mentioned in chapter 2, there are currently three dominant technology families for the deployment of broadband
Internet connectivity:
41
1
2
3
Broadband wireline networks – DSL, cable TV and
fibre solutions,
Broadband wireless access – 3G mobile and WiMAX,
and
Non-terrestrial wireless options – VSAT and other satellite and stratospheric platforms.
Each of these solutions is in play today, and each has
niches in which it is the most appropriate. In this section we
will consider various contexts – geographic, economic, demographic, and public policy environments – and how they might
determine the viability of each family of network solutions.
The starting point of this analysis is that there is no significant
environment on the planet today in which broadband Internet
does not make commercial, social, and institutional sense.
These families of technologies under discussion in this
chapter mostly describe the edge network, that is, the “last mile”
component of the overall network that connects the base station or central office to the subscriber’s premises. The backhaul
network is the facility that connects the various base stations
together. And finally, this traffic is aggregated by the backhaul
network and passed on to the Internet “cloud” which we will
refer to as the core network.
This chapter has focused mostly on edge (or access) networks because, in particular, they are the most expensive and
difficult to deploy. But this does not hide the fact that for
broadband wire-line networks, the availability of core network
facilities, and in particular, fibre networks, is vital to promoting
broadband deployment.
The following subsections explain the four main environments in which edge networks can be deployed: (1) converged
environments, (2) complementary environments, (3) competing environments and (4) exclusive environments. Different
policy and regulatory choices may apply to each environment,
and new regulatory frameworks may have to be developed as
broadband markets evolve and mature.
3.4.1 Converged Environments
Converged environments are found in densely populated areas
with high-value fixed and mobile service subscribers, such as
in metropolitan areas in North America. In a converged environment, fixed wire-line networks compete with or complement BWA networks. And BWA networks are being integrated
so that users of converged appliances (mobile handsets, personal digital assistants (PDAs) or laptops) can move seamlessly
between 2.5G or 3G networks and WiFi or WiMAX networks.
As these converged environments develop, two approaches
to the integration of the wireless networks are being explored.
Tightly coupled integration can be implemented at a relatively
low level within the network. For example, a WiMAX network can be made to appear to a 3G mobile network facility as
just another cellular access network. So there can be seamless
handoffs between cellular and WiMAX networks.
By comparison, another approach, loose coupling, integrates networks at the Internet Protocol layer. This form of
integration will still support limited handoff between networks.
Finally, and more immediately, hybrid appliances can support a
42
limited form of network convergence, for instance today’s WiFi and 2.5G-enabled handsets.
Converged environments are the most capital- and technology-intensive, requiring significant vertical integration
within network operators. They also place significant cost burdens on each user, because converged (or multi-mode) handsets and appliances are only now becoming available on the
market and are consequently expensive.
3.4.2 Complementary Environments
Converged environments are just now emerging. In
today’s reality, most broadband operators either compete or
complement each other (or both). In a complementary environment, broadband network providers offer their services to specific niche markets. Networks are often broadly pervasive but
not wholly ubiquitous. For instance, a single user may subscribe to three different broadband network services provided
by three separate operators:
1 A cable-based home broadband network,
2 A WLAN hotspot service available at coffee shop chains,
airport terminals, and other public places, and
3 A 2.5G EDGE network provided by a major cell phone
operator.
None of these networks currently are integrated, as they
would be in a converged environment. But they all respond
to specific niche demands, often from the same end user. The
same individual would use different services when working at
home, enjoying a cup of coffee or tapping into email while on
the go.
3.4.3 Competing Environments
There are many areas where the number of high-value
subscribers is substantial, but the organizational and capital
environment has not yet been able to support much in the way
of converged facilities. In these areas, multiple broadband networks will be deployed in competition. Indeed, this is the case
for most metropolitan areas in Europe, North America and
parts of Asia and Latin America.
Even an area such as the complementary environment
described in the previous subsection, with three complementary broadband networks, there will also be competing network
facilities. For example, the dominant local cable TV provider
will offer home broadband (cable modem) service. Competing with this service might be a DSL home broadband service
provided by the incumbent local telephone operator. Similarly,
multiple WLAN hotspot services might compete with each
other, although not often at the same physical location (a single
provider will often have exclusive rights in any given hotspot).
Indeed, we expect that in areas that enjoy a large number
and high density of broadband users, instances of converged,
complementary and competing broadband networks may exist
in a simultaneous and overlapping manner.
3.4.4 Exclusive Environments
In marked contrast, much of the developing world has
no broadband provider at all. While regulators seek to pro-
C HAPTER 3
mote competitive broadband provision, a more likely scenario,
at least in the short term, may be broadband provision in an
“exclusive” environment. It is not that regulators seek to create
monopoly markets. The problem is that they may only succeed,
at least in the short term, in attracting one broadband provider.
Sparsely populated rural areas, low-income areas, and
places with challenging physical environments are all candidates for such exclusive broadband providers. These environments are the most important and the most compelling for
bridging the Digital Divide. Since only a single broadband
network is likely to be commercially viable, at least initially, it
is critical for potential broadband providers to pick the technology family and governments to put into place the institutional
and policy environment to support the network.
How can we best conceptualize and support broadband
services in these exclusive environments, especially in those
“green-field” areas where there currently is no broadband network access? It is useful to consider the three technological
platform families separately. The following subsections discuss
the two “extreme” cases: fixed wire-line access and non-terrestrial wireless access, followed by the still-challenging case for
fixed and mobile terrestrial wireless networks.
3.4.4.2
Wire-line Broadband for Exclusive Environments
Wire-line broadband platforms make good sense in areas
with a large population of stationary users, and where operators have sufficient access to capital, public policy support, and
organizational integration. Indeed, fixed-wire networks are
the cheapest per bit, when the goal is to transmit a very large
numbers of bits. Researchers have argued that fixed wire-line
solutions are the most cost-effective, compared with available
wireless solutions, in environments with more than 40 broadband subscribers per square kilometre.
The choice between the three primary wire-line technologies (DSL, cable, and fibre) will probably be driven mostly by
the predominant physical layer of the existing networks. If there
is a high-quality copper network already in place, then broadband DSL service might be the easiest and most cost-effective
to implement. Similarly, if cable TV systems pass most homes,
then those systems’ coaxial and fibre networks might be able to
support broadband data.
It is important to note, however, that exclusive environments are usually in areas that have neither a large population of high-value users nor the necessary capital to support
the financially intensive requirements of building broadband,
wired networks. In other words, few if any exclusive environments would be best served by a fixed wire-line broadband
network.
3.4.4.2
Non-Terrestrial Platforms for Exclusive Environments
VSATs are the clear solution for extremely remote areas
and areas with extreme geographic conditions (for example, mountainous regions). Today, every part of the planet
is touched by multiple satellite signals providing, for a price,
broadband Internet services. But high prices tend to make
satellite connectivity suitable largely for the most remote and
difficult settings (where there are no alternatives). It is not
C HAPTER 3
uncommon to be confronted with a monthly fee of thousands
of dollars for broadband satellite service.
One possible approach that leverages economies of scale
and creates opportunities for sharing expenses is to utilize
VSAT technologies for backhaul networks. A local BWA network can then share this connectivity with multiple local subscribers.
3.4.4.3
Terrestrial Wireless Networks for Exclusive
Environments
It is likely that the majority of exclusive environments,
especially in rural and semi-urban areas of low- and middleincome countries, will receive broadband Internet access using
some form of terrestrial wireless technology. This subsection
considers the case for WiMAX and related networks in these
contexts, exploring the business and policy environments that
are needed to support these networks, then contrasting these
networks with their mobile 3G counterparts.
WiMAX, Wi-Fi, 802.20 and related networks are an attractive solution in many environments, especially if there are likely
to be five or fewer subscribers per square kilometre. Deployments can be bottom-up and potentially incremental. Wireless
systems can work in environments with relatively weak institutional support and small capital markets.
Exclusive BWA networks can be visualized as a series of
circles connected by sticks. The circles represent areas of coverage by a point-to-multipoint radio transceiver, representing
the edge (access) network. The sticks represent the backhaul
network that ultimately aggregates the traffic at a head end
and passes it on to a higher-tier ISP – the core network. The
“circle” of edge connectivity need only have a radius of hundreds of metres, using popular Wi-Fi systems. The backhaul
networks can have “sticks” that extend for many tens of kilometres, using WiMAX-related technologies.
Alternatively, the edge-network circles could have a radius
of 10-20 kilometres, employing WiMAX base stations and
omni-directional antenna facilities. They, in turn, could be
backhauled with long-distance, point-to-point microwave networks or fibre, wherever they exist.
In an exclusive environment with a very small number
of anticipated subscribers, the fixed-wireless approaches of
WiMAX and related networks promise to be the most costeffective approaches. One reason is that they can connect only
those specific areas requiring service, rather than spreading a
ubiquitous signal over a wide area. This is a benefit in sparse
and low-demand areas. Furthermore, most WiMAX technologies do not support vehicular speeds (though newer 802.20 and
WiMAX technologies are emerging that do handle mobility),
nor do they usually support seamless handoff between cells.
As noted in Figure 3.4 above, however, distinctions in
mobility, cost, ubiquity, and even capacity between WiMAX,
802.20, and the 3G family of networks, are eroding in favour of
rapid convergence on a fairly similar feature set. For instance,
mesh Wi-Fi and WiMAX systems are available that can provide
nearly seamless and ubiquitous coverage. Similarly, WiMAX
43
Figure 3.8: Mobility v. Data Rate for Popular BWA Systems
standards (and 802.20) that support vehicular motion are in the
works.
3.4.4.4
3G for Exclusive Environments
Mobility and ubiquity are key elements of today’s 3G
deployments. These may not, however, be critical for exclusive environments. In under-served and remote environments,
a ubiquitous and mobile broadband connection, with seamless handoff at vehicular speeds, may be more than the market
needs or can bear. In fact, there are many examples of deployed
2G networks that are used mostly as fixed-mobile voice networks. The Grameen Phone network is well known for providing rural voice services in Bangladesh via village “phone ladies.”
Providing support for vehicular travel and seamless handoff
to these “phone ladies” is, according to the project’s principal
organizer, an engineering solution in search of a problem. And
if vehicular mobility is generally not needed for voice traffic, it
is questionable whether it is necessary for broadband data.
In some areas, there may well be sufficient markets for
ubiquitous and mobile 3G in exclusive environments, including relatively remote and sparsely populated areas. After all,
mobile 2G networks enjoy very high penetration, even in
some of the most remote and low-income areas in the world.
One potentially viable approach may be to leverage existing 2G
networks (or even build new, green-field mobile networks),
upgrading them to 2.5G or 3G support. This may well depend
on industry response and the regulatory environment.
To consumers, 3G and fixed BWA networks may appear
to be converging, as fixed wireless adds mobility and ubiquity
and 3G increases capacity. The two technologies do, however,
come from very different communities. Fixed wireless tech-
44
nologies have a data networking and information technology
pedigree, while 3G services are primarily telecommunication
industry products. This difference is reflected in management
of the two different kinds of networks. 3G networks are usually deployed after top-down planning, by large, vertically integrated operators with access to substantial investment capital.
Corporate headquarters announce new deployments, which
commonly cover large areas. Fixed wireless networks, by contrast, have often been driven by communities, small entrepreneurs and retailers, who employ an incremental approach.
There are many tradeoffs between the various types of
BWA networks that can be deployed within an exclusive environment. In some cases, upgrading 2G networks to 3G capabilities will be the most promising solution. In other areas, the
2G base station infrastructure might be suitable for deploying
WiMAX (or even 802.20) access points. Alternatively, in some
green-field settings it may be desirable to abandon the mobile
phone market and target specific areas for fixed wireless solutions. In other green-field environments, it may be most
appropriate to leapfrog straight to a 3G mobile network that
can provide broadband data to an entire region.
3.5 Power requirements for broadband
A reliable electrical power source is crucial to achieving
pervasive broadband connectivity. High-capacity wireless networks, and the multimedia appliances that broadband applications call for, can heighten electric power demand. So any
effort to deploy broadband networks must also consider the
availability of power solutions, both on and off the electrical
C HAPTER 3
Figure 3.9: Power Consumption of Some Wireless Radio Products
grid. Today’s modern desktop personal computers (PCs) need
electricity to power:
–
The central processor (now often consuming 80 watts),
–
The screen or monitor (consuming up to 100 watts),
–
The hard drive (which can consume 100 watts when spinning),
–
RAM memory, graphics cards and other on-board devices
(up to 40 watts).
Low-power versions of these technologies do exist today.
For instance, power-efficient laptops can be engineered to
require less than 40 watts under normal operation.
Figure 3.9 shows the power consumption for a set of
popular WLAN radio technologies. “BreezeMAX” WiMAX
technologies can require 30-40 watts when in operation. A
subscriber unit for Motorola Canopy, on the other hand, consumes less than 10 watts, while an indoor wireless router from
Links consumes slightly more than 10 watts, on average.
What these figures suggest is that a standard, broadbandenabled PC can consume a fairly considerable amount of
power; even low-power laptops require a significant amount.
Furthermore, the power load of a PC is non-linear; it fluctuates over time (unlike a light bulb, for example), and a PC
requires high-quality power sources.
This leads to the question of whether high-quality, reliable power sources are available in areas that have inadequate
or unreliable access to the electrical grid. Major solutions for
off-grid powering of ICT systems include solar cells, small
wind systems, micro-hydro power technologies, and generator
sets (powered by diesel, gasoline, or bio-fuels). Each of these
solutions has its benefits, as well as drawbacks. The following sections provide two case studies that illustrate innovative
C HAPTER 3
applications of off-grid power for ICTs in developing countries.
3.5.1 Tanzania: The Kasulu Teachers Training
College
At the Kasulu Teachers Training College (KTTC) in Tanzania, the “First Wave” of agrarian reform has mingle with the
wave of new information technologies to create a broadband
service that is changing the lives of some of the most impoverished people in Africa. An Internet centre at the KTTC equips
the next generation of teachers in Tanzania with Information
Society tools, despite the fact that Kasulu has no electricity
supply and its 15 phone lines have no data capabilities. What is
unique about the project is that it is run by an unusual source
of power – cow manure.11
Kasulu, population 33,668,12 is located in northwestern
Tanzania. Being close to the Burundi border, the region has
seen a huge influx of refugees escaping from violence in that
country. In 2000, the Global Catalyst Foundation (GCF), in
collaboration with the United Nations High Commissioner
for Refugees (UNHCR) and Schools Online, decided to
explore the feasibility of setting up a Kasulu Internet Project13
to promote cooperation and understanding between refugee
Burundians and their Tanzanian neighbours in the Kasulu area.
Another objective was to promote economic development and
entrepreneurship in the impoverished region.
3.5.1.1
The Kasulu Programme
GCF is a private foundation established by Kamran Elahian, an IT entrepreneur and founder of Global Catalyst
Partners, a Silicon Valley venture capital firm. GCF’s mission
is “empowering people through technology,” and it supports
45
Figure 3.10: UNHCR Refugee Camps in Tanzania
Source:
Food and Agriculture Organisation.
projects that “improve education, alleviate poverty, promote
social tolerance and celebrate diversity” across the world.14
Schools Online, another non-profit started by Elahian, aims to
help students use the Internet for learning and cross-cultural
dialogue. Since 1996, the organization has assisted more than
5,700 schools in the United States and more than 400 schools
in 35 other countries, helping to procure the equipment and
support services to get online.15
Through the Kasulu Internet Project, GCF, Schools
Online and UNHCR have collaborated in establishing Internet centres in three locations:
–
At a UNHCR administered refugee camp, dubbed “Meatball,”
–
At the Kasulu Folk Development College, and
–
At the Kasulu Teacher Training College.
GCF contributed USD 120,000 in operational funds for
the three centres for three years; it is responsible for the overall
management of the project. The local communities have contributed their labour to build computer labs. UNHCR provides logistical and administrative support, and Schools Online
provides the hardware and software infrastructure and funding
for professional development and capacity building.16
At the KTTC facility, the 800 students enrolled in the
teacher-training programme are the main users of the computer labs, which are open to community members from 7-10
pm during the week (they are also open on Saturdays and
Sundays), on a fee-for-use basis. The price is about USD 1 per
hour (1,000 Tanzanian Shillings). Local representatives of non-
46
governmental organizations and UNHCR are frequent users
of the labs, as well.
The teachers and students take CISCO Academy courses
and study for the International Computer Driver’s License
(ICDL), which has seven skill levels in word processing, file
management, Internet services and other fields. The CISCO
Academy has seven students at KTTC, who pay approximately
USD 200 for a six-month course. Some 10 students from the
town of Kasulu are also studying elements of the ICDL curriculum, paying approximately USD 200 per course. Another
30 students from the KTTC also attend these classes but pay
lower fees. Revenues from these training programmes help
KTTC pay for the cost of Internet access, maintenance and
ongoing operations.
The Tanzanian Ministry of Education is now planning to
make IT training compulsory for teachers. This will increase
the demand for computer classes at KTTC, because all of the
area’s teacher candidates will require instruction.
3.5.1.2
Kasulu’s Internet Links
The school is connected to the Internet through a VSAT
service provided by I-way. The I-way connection offers
128 kbit/s downlink and 30–40 kbit/s uplink speeds. The Global
Catalyst Foundation pays USD 500 per month for the connection. The Internet is reasonably reliable, although it does fail
occasionally.
As mentioned above, the GCF turned to an unusual
source of power – cow manure – to power the KTTC compu-
C HAPTER 3
Figure 3.11: The KTTC Computer Laboratory
Source:
Michael L. Best
Figure 3.12: KTTC’s Eco-Friendly Power Plant
The biogas digester (top left), methane/diesel generator (bottom left), and cow corral (right).
Source: Michael L. Best
ter lab. The droppings of twelve cows are collected and fed into
a 50-cubic-metre biogas plant that generates methane. This
methane is then mixed with diesel in a 70-30 ratio and fed into
a power generator producing 10 kilowatts of power – enough
to run 15–16 computers for eight hours daily. Six shared UPS
systems provide a 30-minute power backup.
Nothing goes to waste in this system. After the methane
is extracted, the remaining sludge is removed from the biogas
plant to provide fertilizer for the crops the college raises to
feed its staff and students. The biogas processor, the dairy
cattle, cowshed and 10-kilowatt generator cost a total of about
C HAPTER 3
USD 18,120. The biogas system was built in late 2001, and the
computers arrived at KTTC 2002.
Apart from powering computers, the college also wants
to use methane for cooking. Currently, the college consumes
several tons of timber every year to cook food for its students.
Replacing timber with methane would be more ecologically
friendly, and the cooks would also be spared the harmful effects
of wood smoke. To do this, the college plans to increase the
capacity of the biogas plant from 50 cubic metres to 200 cubic
metres. To provide manure for the increased capacity, the college would have to maintain 40–50 cows. The additional meth-
47
Figure 3.13: The Mtabila Camp VSAT and PV Systems
Mtabila camp VSAT and PV systems (top); refugees waiting outside the computer lab (bottom left); and inside the computer lab (bottom right).
Source:
Michael L. Best
ane generated would fuel a 30-kilowatt generator and allow for
a 100-per cent methane-powered system.
The Swedish International Development Agency (SIDA)
is considering funding the expansion. It would like to replicate the biogas system at 30 teacher training schools in Tanzania and perhaps 3,000 secondary schools. The organization
is also funding a micro-hydro power project in Kasulu where,
along with a power micro-grid, fibre optic cables will also be
strung. The fibre optic cables will be connected to the Internet
through VSATs and used for networking NGOs, schools and
clinics.
3.5.2 Tanzania: The Mtabila Refugee Camp
More than 50,000 Burundian refugees stay at the Mtabila
refugee camp, where conditions at the camp are bleak and there
is very little work to occupy the residents. The refugees often
live in simple houses and survive mostly on food provided by
the World Food Programme (WFP). Most of the adults at the
camp are women,17 although the camp also has a significant
adolescent population that faces an uncertain future.
With no telephone service in the camp, the 10 computers
installed at the Mtabila Internet Centre are the refugees’ main
connection to the outside world. When the centre is open to
the general public there is often a long wait for Internet access.
Every day, a clientele of about 30 refugees use the centre to
send email, at a charge of approximately USD 0.20 (20 cents)
per session.18 The Internet has reduced the refugees’ sense of
isolation and enabled them to communicate with the outside
world at a cost that is cheaper than the postal service. News
and information downloaded from the web also help the refu-
48
gees stay abreast of world events. The refugees also learn to
access websites, including those in the Kirundi and Kiswahili
languages.
The Internet Centre is equipping refugees with skills they
can use to help reconstruct Burundi. The Kasulu Online19
website reports that some 2,000 refugees have received higher
education, directly benefiting from Internet access. Other refugees benefit through intermediated access, as others pass on
current events to them and help them establish contact with
friends and relatives who have email.
Ten teachers, selected by the refugee community, are
undergoing training for the International Computer Driving License (ICDL) at the camp. These same individuals also
supervise the centre. The big surprise for these teachers, as well
as camp administrators, has been the demand for email. Many
of the refugees have friends and family scattered throughout
Africa, Europe and North America, and email is an affordable
way to keep in touch with their community.20
Broadband Internet access is provided through a VSAT
terminal, with the electrical power generated using photovoltaic cells. The system’s setup cost approximately USD
37,152. It consists of 48 solar panels, generating 75 watts each.
Solar power was chosen over biogas in this case, because solar
eliminated dependence on fuel supplies and spare parts that a
generator-powered system would entail.
Setting up the centres was a demanding task because of
the poor quality of the telephone lines, which could barely support email transmissions. Also, there were delays in importing
essential equipment and challenges in dealing with relations
among various governmental, international and non-govern-
C HAPTER 3
mental agencies involved in the project. Despite these hurdles,
the project was beginning to deliver the intended benefits, as
can be seen from a refugee’s email from the Mtabila camp:
“I’m not able to tell you how happy we are to get connection
to the Internet! Before I was connected to the Internet I felt
lost. But now that I am connected, I feel saved. The world will
not forget us now, because we, the refugees, can speak to the
outside.”21
3.6 Conclusion
This chapter has surveyed the development paths of three
important families of broadband access technologies: wire-line
services, broadband wireless access, and non-terrestrial broadband wireless systems. All three of these families of systems can
provide optimal choices in various circumstances. In densely
populated, high-income areas, a variety of these systems will
converge, compete, or complement each other. In other settings – those with limited numbers of potential broadband
subscribers – single network systems are likely to prevail in
isolation. And in cases where the density of potential subscribers is lowest, various wireless broadband solutions are likely to
be the most cost-effective.
These sets of technologies can be placed on a continuum,
based on the density of high-value subscribers. With densities
of 40 people per square kilometre and above, wire-line solutions are likely to be cost-effective. As density declines to just a
few people per square kilometre, fixed wireless solutions, such
as WiMAX, are more optimal.
Broadband deployment does not depend solely on technological choices. Capacity building to develop local human
C HAPTER 3
resources is vital. Education and training – including practical
showcases and pilot programmes – are equally important in the
quest to promote broadband access. Governments can further
organize open regional, national and local forums dedicated to
identifying and meeting the broadband needs of all key stakeholders, including:
•
End users (both current and potential);
•
National and local development programmes on healthcare, education and e-government;
•
Public sector institutions such as universities, libraries and
local and national government offices;
•
Local entrepreneurs;
•
Traditional telecommunication operators;
•
System manufacturers; and
•
Owners of fibre infrastructure from other sectors, such as
transport and utility companies.
Together, these stakeholders can strive to meet their broadband requirements and achieve the UN Millennium Development goals. Once broadband needs are identified, they can be
met through the development of open market-based networks,
fostered through effective regulatory practices.
Broadband deployment in rural areas of developing countries is likely to be fuelled both by civil society, including a range
of public service actors, and the private sector, including local
entrepreneurs that are building sustainable businesses. What is
needed are regulatory frameworks designed to lower business
risks and open markets to a full range of potential broadband
providers. These key issues are addressed in the next chapter,
which discusses the role of regulators in providing an environment that promotes broadband deployment.
49
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
See http://standards.ieee.org/getieee802/portfolio.html for more information about the IEEE802 standard.
In Sweden, apartment house companies own their own fibre networks to serve the tenants in their buildings. These networks connect to a point of presence of all
operators at a metropolitan area hub. In a departure from the fibre to the home model, they outsource the link level of their fibre from the home networks to serviceproviders serving their tenants.
See: http://www.itu.int/ITU-D/treg/publications/AfricaIXPRep.pdf
See http://grouper.ieee.org/groups/bop/ for further information about P1675 from the IEEE).
See http://www.iec.ch/tctools/dashbd-e.htm for further information about the working groups with the CISPR.
For further information about this process, please see http://www.iec.org/online/tutorials/acrobat/opt_net.pdf.
See http://www.iec.org/online/tutorials/sdh/ for more information about SONET.
Commercial acceptance has been disappointing on these systems and it is thought that ArrayComm intends to abandon this technology in order to focus on WiMAX
based systems.
ITU Internet Reports: The Portable Internet, 2004, see http://www.itu.int/publications/bookshop
Portable Internet.
“Cow pats fuel computers” BBC Online. See http://news.bbc.co.uk/2/hi/technology/2957488.stm
See http://www.tanzania.go.tz/census/districts/kasulu.htm
See http://www.kasuluonline.or.tz/
See http://www.global-catalyst.org/
See http://www.schoolsonline.org/whoweare/index.htm
See http://www.schoolsonline.org/whatwedo/update_sep_2002.htm
“Rape at the end of the world.” See http://www.unfpa.org/focus/tanzania/rape.htm
“Cow pats fuel computers” BBC Online. See http://news.bbc.co.uk/2/hi/technology/2957488.stm
See http://www.kasuluonline.or.tz/
“Tanzanian refugee camp gets wired for Internet” See http://hrea.org/lists/huridocs-tech/markup/msg00932.html
“Start. Succeed (or Not). Repeat.” Los Angeles Times. See http://www.latimes.com
50
C HAPTER 3
4 THE ROLE OF THE REGULATOR IN BROADBAND
DEVELOPMENT
Authors: Yang-Soon Lee, Dr. William Bratton & Wu Wei Shi, Spectrum Strategy Consultants and Susan Schorr, ITU/BDT
4.1 Introduction
Previous chapters have explored the promising landscape
of new broadband opportunities that is emerging as waves
of innovation reshape the ICT sector. These innovations are
occurring across all aspects of the industry, from technological developments and business models to regulatory and policy
frameworks, creating broadband opportunities for end users,
large-scale network operators, small entrepreneurs, local communities and governments alike.1 New broadband opportunities require a new vision by potential broadband providers, and
a new paradigm for policy-makers and regulators. Broadband
is completely transforming the ICT sector. Put simply, broadband cannot be treated as “business as usual.”
Network operators that fail to join the broadband world
risk being left behind. The reality is, however, that deployment
of broadband access technologies in developing countries is
often constrained by a lack of telecommunication infrastructure – especially backbone networks – and by large-scale network operators’ concerns about potential revenue generation.
Larger commercial operators are often discouraged from providing broadband access in marginal areas because of the high
costs of deploying networks and the fear that retail charges
may be too high, relative to disposable incomes, to result in
large-scale take-up.2 Meanwhile, potential broadband market
entrants are often kept out of the market by regulatory frameworks designed for another era.
A new, pro-broadband regulatory paradigm will harness
the power of all potential broadband providers, tailoring the
regulatory framework as needed. This pro-broadband regulatory paradigm will expand on the existing regulatory practices
developed after the revolutionary advent of the internet. Broadband technologies now allow discussions about the internet to
be taken to a higher level, because they accelerate download
speeds and provide extended mobility. In response, regulators
will seek to spur competition at all levels of the broadband
value chain, from the link and transport infrastructure layers
to the content required to fuel demand and the computers
needed to access broadband services and applications. This will
require regulators to take a comprehensive and coordinated
approach, as identified by the world community of regulators
participating in the 2004 ITU Global Symposium for Regulators (See Box 4.1).
C HAPTER 4
The new broadband regulatory paradigm will paradoxically
require regulators to do both more and less than “regulation as
usual.” Regulators will do more in making potential broadband
providers – such as local communities and non-governmental
organizations – aware of the technologies and applications they
could provide. Regulators will also coordinate more actively
with other government agencies and public institutions (such
as universities) to drive the demand for health, education and
government services that employ broadband technologies and
applications. Meanwhile, regulators will do less by dismantling
outdated regulatory frameworks that restrict market participation.
A successful, comprehensive policy framework in a developing country with major challenges in rural connectivity can
not only foster greater investment by large scale-operators, it
can encourage public institutions and smaller market players
to deploy broadband networks to suit their own operational
or commercial objectives. As highlighted in Chapters 2 and 3,
many broadband technologies can be deployed incrementally,
without large-scale, nationwide deployment plans. The ability to deploy broadband on an incremental basis allows local
communities and development project managers to include
broadband capabilities in multiple projects. Similarly, small
and micro entrepreneurs can launch new businesses based on
broadband access.
The key is for regulators to determine how to minimize
any obstacles to incremental deployment, which may include:
• Prohibitive pricing for interconnection with incumbent
operators;
• High costs to access existing infrastructure or resources
from parallel utility sectors (i.e., energy or transport utilities); and
• Extensive and onerous licensing processes.
Once small-scale market entrants can establish demand for
broadband services, larger operators are more likely to see the
benefits of market entry, themselves. If need be, their interest
can be further awakened through appropriate rewards schemes,
such as tax exemptions or targeted subsidies from universal
service funds.
It is also important to highlight programmes and content
that may drive initial broadband take-up in areas where it has
51
Box 4.1: GSR 2004 Best Practice Guidelines
Regulators meeting at the 2004 Global Symposium for Regulators, held 8-10 December, 2004 in Geneva, approved a set of
“best practice” guidelines for regulators to promote the development of low-cost, easily accessible broadband networks and applications. The full text of the document is as follows:
“We, the regulators participating in the 2004 Global Symposium for Regulators, have identified and proposed best practice
guidelines to achieve low cost broadband and internet connectivity. Our goal is the creation of national regulatory frameworks
that are flexible and enable competition between various service providers using multiple transport and technology options. We
believe the best practices outlined below will help bring social and economic benefits to the world’s citizens.
An enabling regulatory regime that encourages broadband deployment and internet connectivity
1)
We encourage political support at the highest government levels with such support expressed in national or regional policy
goals. These include an effective regulator separated from the operator and insulated from political interference, a transparent regulatory process, and adoption and enforcement of clear rules.
2)
We believe that competition in as many areas of the value chain as possible provides the strongest basis for ensuring maximum innovation in products and prices and for driving efficiency.
3)
We encourage regulators to set policies to stimulate competition among various technologies and industry segments that
will lead to the development and deployment of broadband capacity. This includes addressing barriers or bottlenecks that
may exist with regard to access to essential facilities on a non-discriminatory basis.
4)
We believe that the primary objective of regulation should be to secure fair and reasonable access for competitive broadband
services, including internet connectivity.
5)
We encourage the maintenance of transparent, non-discriminatory market policies in order to attract investment.
6)
We encourage regulators to adopt policies that are technology neutral and do not favor one technology over another.
7)
We encourage regulators to take into consideration the convergence of platforms and services and that they regularly reassess
regulatory regimes to ensure consistency and to eliminate unfair market advantages or unnecessary regulatory burdens.
8)
We encourage regulators to allocate adequate spectrum to facilitate the use of modern, cost effective broadband radiocommunications technologies. We further encourage innovative approaches to managing the spectrum resource such as the
ability to share spectrum or allocating on a license-exempt non-interference basis.
9)
We urge regulators to conduct periodic public consultations with stakeholders to inform the regulatory decision-making
process.
10) We recommend that regulators carefully consider how to minimize licensing hurdles.
11) We encourage the development of a regulatory framework that permits ISPs and broadband providers to set up their own
last mile.
12) We encourage regulators to provide a clear regulatory strategy for the private sector in order to reduce uncertainty and risk,
and remove any disincentives to investment.
Innovative Regulatory Policies Must Be Developed To Promote Universal Access
1)
We recommend that the promotion of access to low cost broadband interconnectivity should be integrated from “grassroots” efforts to identify local needs all the way through the “tree-tops” of international law. Governments, business and
non-governmental organizations should be involved.
2)
We recommend that regulators adopt regulatory frameworks that support applications such as e-education and e-government.
3)
We encourage each country to adopt policies to increase access to the internet and broadband services based on their own
market structure and that such policies reflect diversity in culture, language and social interests.
4)
We encourage regulators to work with stakeholders to expand coverage and use of broadband through multi-stakeholder
partnerships. In addition, complementary government initiatives that promote financially sustainable programmes may also
be appropriate, especially in filling in the market gap that may exist in some countries.
5)
We encourage regulators to adopt regulatory regimes that facilitate the use of all transport mechanisms, whether wireline,
power line, cable, wireless, including wi-fi, or satellite.
6)
We encourage regulators to explore programmes that encourage public access to broadband and internet services to schools,
libraries and other community centres.
7)
We encourage regulators to implement harmonized spectrum allocations consistent with the outcome of ITU Radiocommunication Conference process and each country’s national interest. Participation in this well-established framework will
facilitate low-cost deployment of equipment internationally and promote low-cost broadband and internet connectivity
through economies of scale and competition among broadband vendors and service providers.
52
C HAPTER 4
Box 4.1: GSR 2004 Best Practice Guidelines
Broadband is an Enabler
1)
Regulation should be directed at improving the long term interests of citizens. Broadband can contribute to this by improving and enabling education, information, and increased efficiency. It can reduce costs, overcome distance, open up markets,
enhance understanding and create employment.
2)
We encourage regulators to educate and inform consumers about the services that are available to them and how to utilize
them so that the entire population benefits.
3)
We urge regulators to work with other government entities, industry, consumer groups, and other stakeholders to ensure
consumers have access to the information they need about broadband and internet services.”
See: http://www.itu.int/ITU-D/treg
been absent. In developing countries, governments are often
the largest ICT users. Governments can launch top-down egovernment initiatives to induce citizens to get online, where
they can avoid bureaucratic appointments, queues and paperwork and save on travel costs. E-government services also help
the government itself by developing ICT capacity to interface
with the citizenry en masse. But e-government applications will
only drive broadband take-up if the public accepts them as
useful.
Once demand is proven, commercial broadband providers
will respond with entertainment content designed and priced
for developing-country users. In short, promoting broadband
access and services in urban and rural areas of developing
countries requires a new vision characterized by reduced regulatory burdens and an environment of innovation, collaboration and creativity.
4.1.1 A Look Back
Ten years ago, the predominant model for providing telecommunications involved a state-owned incumbent operator deploying a fixed-line public switched telephone network
(PSTN). Since then, that model has given way to alternative
competitive network platforms such as cable television, fibre
optic, satellite, second generation (2G) mobile cellular, and
more recently, 3G and broadband wireless – all enabled by
new technologies. Most telecommunication networks have
been provided on a nationwide, or at least regional, basis.3
Today, most large-scale networks are in private hands, meaning they must be profitable to operate. Even the majority of
PSTNs today have been at least partially privatized.4 The business model adopted by large-scale network operators requires
significant infrastructure investment and high subscriber revenues to remain financially viable. Most national regulatory
frameworks were designed for such large-scale network operators and service providers.
What has this first wave of sector reforms achieved? The
short answer is: a lot. Telephone access has quadrupled since
1990, from 10 per cent to 40 per cent of the world’s population. By the end of 2004, there were an estimated 1.19 billion
fixed telephone lines in operation around the world and 1.8
billion mobile subscribers. The rise in mobile subscribers is
phenomenal, and more than 50 per cent of mobile subscribers
today are in developing countries.5 Still, many of these users
C HAPTER 4
are located in urban areas. The challenge for regulators today
is to build on this urban mobile growth to provide an enabling
environment to bring both voice and broadband internet services to rural areas.
The stage is now set for such a rural renaissance. An
entirely new player has appeared on the scene: the community
broadband provider, largely enabled by low-cost broadband
wireless technologies. Local community initiatives are starting to provide broadband services to users in remote areas not
served by large-scale networks. These small-scale broadband
providers range from public institutions such as libraries, educational institutions, health facilities and local governments, to
non-governmental organizations and small and micro entrepreneurs that can be profitable at margins of only a few dollars a day. Whether private or public, all community broadband
providers are concerned with costs. The lower the costs of
providing broadband services and applications, the greater the
opportunity for community initiatives to succeed in bringing
ICTs to rural users. Keeping regulatory costs down will give
these local initiatives a better chance of success.
4.1.2 A Look Ahead
A new regulatory framework, tailored to the unique circumstances of local community initiatives, may be needed to
help small-scale broadband providers foster growth in rural
areas. This chapter explores options for such a new regulatory
framework.
Meanwhile, the regulatory framework designed for largescale network operators is also evolving to provide greater
incentives to deploy broadband access networks in rural areas.
This chapter takes a hard-nosed look at the commercial realities
of major operators. Commercial operators are profit-driven and
cannot be expected to provide services that do not yield profits.
Where regulators are unable to induce large operators to build
broadband networks in rural areas, either through financial
incentives or build-out requirements, they can focus efforts on
developing new pools of potential broadband users. This new
source of demand will then give major operators incentives to
extend their networks out to these new markets, offering internet backhaul and interconnection with urban networks.
This chapter first identifies the importance of broadband
and the potential positive impacts that access to broadband
technologies can provide, before discussing the key issues
53
associated with such deployment in developing countries. The
chapter then highlights the role of competition in accelerating
the deployment of supporting networks and reaches some conclusions about the appropriate regulatory framework to create
a conducive environment for deploying broadband networks.
The chapter will conclude by identifying other potential policy
measures regulators can undertake to support the deployment
and take-up of broadband access technologies.
4.2 The Importance of Broadband in Developing
Countries
The deployment of broadband access technologies delivers several positive impacts in developing countries. These
include:
• Eroding information differentials resulting from geographical constraints that prevent marginal communities
from participating in regional, national or international
processes.
These processes may be social (for example, relating to
education or health), political, economic or financial. Residents of rural areas, for example, need access to financial information and advice, as well as information about markets (both
as buyers and sellers). Broadband technologies, therefore, can
facilitate the integration of marginal communities into wider
processes beyond the geographical limitations of their specific
areas.
• Access to regional, national and international
resources through broadband access technologies can
substantially improve the living standards of marginal
communities.
For example, improving accessibility to e-health systems
that allow remote diagnoses and treatment is particularly
useful in marginal areas, where access to medical equipment
and expertise would otherwise be limited. Similarly, broadband
access technologies can be used to provide remote education
and training services (for example, the African Virtual University6 and, at a more advanced level, Universitas 21).7 The provision of such access can have a significant positive impact on
living standards.
• Broadband access technologies can enhance the sustainability of marginal communities by supporting the transfer of knowledge and expertise to marginal communities.
Rural doctors can receive regular training via e-health systems, allowing them to stay in their communities rather than
leaving, either temporarily (for training) or permanently (for
professional growth and improved fortunes). Online training
also brings to the community the latest medical techniques and
treatment options. Similarly, broadband access technologies
can enable the transfer of agricultural knowledge, improving
productivity and limiting soil exhaustion and desiccation from
unsustainable farming techniques.
54
•
Improved information flows from broadband access technologies can increase the range of market opportunities available to marginal communities.
Farmers need not be limited to local buyers. They can
use broadband technologies to access geographically remote
markets, including auctions. This benefit is particularly pronounced for perishable agricultural products where the use of
online markets lets producers establish contacts with a wider
range of potential buyers across larger geographies, boosting
prices and maximizing rural incomes;
• Broadband technologies can improve business productivity in developing countries.
Access to greater information sources, e-mail and other
supported services (VoIP, for example) allow businesses to
lower their business costs and improve their revenue-generating potential. In India, for example, widespread broadband
deployment is expected to increase labour productivity by 11
per cent, leading to direct employment of 1.8 million and total
employment of 62 million by 2020.8
• The deployment of broadband access technologies may
support the growth of regional and national IT
industries.
This growth can impact positively on GDP growth. For
instance, in Korea (Rep.), broadband deployment has significantly underpinned the IT sector, which accounted for
approximately 50 per cent of the GDP growth rate in 2002.9
The Confederation of Indian Industry’s National Broadband
Economy Committee estimates that broadband will contribute USD 90 billion to the Indian economy between 2010 and
2020.10 Specific examples of how the deployment and take-up
of broadband access technologies has benefited marginal communities and developing countries are illustrated in Box 4.2.
There is no doubt that broadband access technologies can
provide substantial benefits to end users in developing countries and marginal communities. But the services provided over
the networks must be sufficiently targeted to offer end users
real benefits. In addition to regulations that aim to create a
competitive broadband market providing universal and equitable broadband access, governments are also focusing on developing policies to include local communities in the design and
implementation of broadband initiatives.
4.3 Key Issues in Promoting Broadband in Developing
Countries
Given the potential economic and social benefits of broadband access technologies, there is a clear need to develop ways
to boost the relatively low levels of broadband take-up in many
developing countries. The fundamental problem is that there
is an array of constraints to take-up in these countries, and they
affect all the components of the value chain (See Figure 4.1).
There are impediments in the supply of online content, as well
as the supply of broadband services and products. There are
also barriers to widespread connectivity. Finally, there currently
is limited demand for broadband in many developing countries.
C HAPTER 4
Box 4.2: The Impact of Broadband Access – Examples in Developing Countries
In the Reserva Ecologica do Xixuanú of Brazil’s Amazon region, a telemedicine project has been launched to transmit medical information from local communities, via satellite, to the United States for remote diagnosis.
Rural villages in Bhutan that were previously not connected by traditional telephone service are now provided with inexpensive basic voice telephone access using wireless broadband technologies.
In China, students in rural villages are able to participate in distance-learning courses set up by major Beijing universities,
using VSAT broadband satellite access.
In a small, remote town in the mountainous region of northeastern Ecuador, Wi-Fi technology has enabled the mayor to
access online government databases. In addition, broadband access is used to promote ecotourism in the area and to help local
businesses (SMEs or “small and medium enterprises”) become more competitive.
In Laos, WLAN networks have been rolled out to villages, allowing people to make local and international VoIP calls, significantly improving their connectivity. Other activities made possible by the WLANs include accounting, letter writing and e-mails,
as well as support of local business activities.
In South Africa, institutions are connected to international institutions using broadband technologies, to advance cooperation in various research and development initiatives.
In Uganda, rural schools can gain access to educational tools via broadband access.
Recognizing the benefits of broadband for the agricultural sector, the government in India has announced plans to set up a
network of computer kiosks in 25,000 villages. The goal is to help farmers sell their produce to the highest-paying customers. The
national rollout is due to be completed in 2007.
Figure 4.1: A Simplified Broadband Value Chain
Limited consumer demand, however, is likely related to a lack
of consumer awareness of the potential benefits of broadband.
Once more broadband networks are deployed, it is quite possible that consumer demand (in terms of content and volume) in
the developing world will not be so different from demand in
the developed world. The main difference may remain in the
rates that consumers are able to pay, and the portion of their
disposable incomes they are willing to allocate to broadband
services and content.
In terms of supply, the key constraints to deployment of
broadband services and products in developing countries are:
• Insufficient compelling content, especially in local languages and with specific reference to local circumstances
or issues;
• Lack of understanding about the benefits of broadband;
• Little or no incentive for fixed-line incumbents to offer
broadband access technologies (particularly if they risk
C HAPTER 4
cannibalizing PSTN and ISDN revenue streams) in the
absence of market competition;
•
Competing demands for investment of operators’ capital
(in developing countries operators often believe they can
generate a better return on capital by deploying mobile
networks than additional fixed-line or broadband networks);
•
Lack of market competition to encourage operators to
develop and commercially deploy broadband services;
and
•
Lack of a regulatory framework designed to encourage
broadband deployment by large-scale incumbents, or to
sponsor market entry by potential broadband competitors,
which might include public institutions (e.g., universities)
or local, community-based providers.
55
In terms of connectivity, the key constraints to take-up
are:
• Lack of hosting/storage facilities within many developing
countries, a situation that requires much content to be
stored overseas, straining international connectivity;
• Limited international connectivity, which impacts on the
data rates available, the quality of the service and the cost
of bandwidth;
• Lack of backbone connectivity in many areas – and where
backbone networks do exist, they are owned by incumbents that control the costs and quality of leased lines
available to competitors; and
• Concern on the part of large-scale network operators over
the commercial viability of deploying broadband networks
in rural or remote areas, where the costs of network operation and service provision may threaten to make services
unaffordable for consumers.
In terms of demand, the key constraints in developing
countries are:
• Lack of consumer demand, resulting from limited consumer purchasing power;
• Lack of consumer awareness (which is linked to pricing
and low purchasing power), coupled with a lack of coordination by key stakeholders (i.e., universities, public
institutions and local communities) that could drive further awareness and demand;
• Excessive pricing of broadband products and services,
especially when compared with average incomes;
• Greater priority placed on mobile voice communications
than data services; and
• Limited availability of affordable end-user terminals.
Deploying broadband access networks only makes sense
if potential users have the computers or handsets with
which to access them.
Many of these factors, of course, are not unique to developing countries. But they are more daunting than in developed countries, where consumers are more likely to be able
to afford broadband services and products at price points that
more accurately reflect the underlying costs of providing them.
Similarly, the relatively low penetration rate of personal computers (PCs) in developing countries significantly weakens
the demand for broadband access. PC penetration is an issue
of affordability, but also a factor of electricity provision, since
it is impractical to use a personal computer where there is no
electricity to power it.
Innovation and competition in developed countries, however, are pushing manufacturers and engineers to develop very
low-cost, simple devices that can be tailored for applications
and content in developing country contexts. This may reflect
the development of a niche market for low-priced, narrowly
functioning equipment, as manufacturers consider how to
cater to untapped markets. Considering the supply, connectivity and demand factors that limit broadband take-up in many
areas of developing countries, regulators have an important
role in minimizing the impact of these factors or developing
appropriate solutions.
56
4.4 Providing Incentives for Network Investment
In developed countries, relatively high levels of teledensity allow for a greater focus on the promotion of service-based
competition. The primary role of most regulators in developing countries, however, is to create an environment conducive
for network investment where little or no telecommunication
infrastructure exists. But before confidently committing themselves to network investment, operators look for government
regulatory approaches that are consistent and not arbitrary. This
is especially true for broadband access technologies, which are
often capital-intensive and – in developing countries – considered a more risky investment.
Therefore, any regulatory framework (and this includes
the track record on enforcement) that creates uncertainty or
the risk of financial loss will deter market entry and subsequent
network investment. Regulatory agencies that fail to provide
consistency risk deterring investment, thereby constraining the
development of their telecommunication networks, including
broadband ones. Within this context, there are several steps a
regulator can take to promote the deployment and take-up of
broadband networks. These are detailed in the following subsections.
4.4.1 New Entry and Market Liberalization
Market liberalization remains the most effective mechanism to encourage greater investment in telecommunication
networks. Experience shows that liberalization through the
licensing or authorization of new operators will yield greater
benefits than incentive- or obligation-driven approaches targeted at only a monopoly or duopoly. The absence of liberalization, meanwhile, removes a significant incentive for an
incumbent to invest in networks, new services or quality of
service (QoS). Historical precedent in countries such as Brazil,
Hungary, and India demonstrates that both incumbents and
new entrants invest more when faced with market competition.
It is clear, however, that after market liberalization, the
incumbents in many developing countries continue to be in
a position to invest significantly more in network deployment than are new entrants, especially in marginal areas. So
the incumbent can often be the most important (and often the
largest) source of funds for telecommunications investment in
the longer term.
Given this, a number of commentators have argued that
excessive market competition may reduce the incentive for
incumbents to continue large-scale investment. This may be
particularly the case where tariff re-balancing has not been
fully undertaken and new entry results in rapid margin erosion for incumbents in the more lucrative long distance and
international markets. This may result in a deterioration in
the incumbent’s financial performance (and stock valuation),
reducing its capability and willingness to undertake new network investment or diversify into new lines of business. Regulatory frameworks, then, should not unduly impact network
investment and the diversity of new lines of potential business.
C HAPTER 4
Box 4.3: Regulatory Methods To Boost Deployment
Regulatory frameworks, tailored to the unique circumstances of local community initiatives, can encourage small-scale
broadband providers to provide broadband access in rural areas.
An incentive-based regulatory framework using targeted grants or tax exemptions can encourage both large and small-scale
network operators to deploy broadband networks in rural areas.
Where financially viable, broadband deployment requirements can be made part of the licensing commitments required
from new market entrants.
Governments can also promote the development of sufficient supporting network infrastructure to enable the provision of
broadband services (e.g. backbone connectivity). They can also drive initiatives (e.g., e-government programmes) that provide an
important source of demand for broadband facilities and services.
Because of the costs of network deployment – especially
of broadband access technologies – there have been some suggestions that network “over-building” (duplication) could be
avoided by creating a “super” network operator that would
provide wholesale network access to retail service providers.
In effect, this would amount to a transition to service-based
competition rather than network-based competition. This idea
often appeals to policy-makers in developing countries, because
of their concerns about the availability of investment funds.
But implementation raises significant competitive concerns.
Incumbent operators tend to advance the argument for
“super operator” status in an attempt to limit competition.
“Super” network status allows incumbents to retain complete
control over infrastructure and, by extension, over the development of competition. In both developing and developed
countries, incumbents have proven remarkably effective at
controlling access to their infrastructure even when the regulator is relatively strong. It is unlikely, therefore, that a “super”
network scheme would be effective in the absence of a very
strong regulator that could ensure that service providers have
equitable access to the network.
One alternative is for governments to fund the construction of such comprehensive networks directly, and to oversee
access to them by both incumbents and new service providers. Such an approach has been used successfully in Singapore,
where the government in 1997 funded the rollout of SINGAPORE ONE, a broadband backbone network accessible by the
incumbent and new operators, including the cable TV operator. Of course, not all governments have the resources to build
such broadband backbone networks – particularly nationwide.
A more practical alternative, therefore, may be offered by
infrastructure sharing. For example, allowing mobile operators
to roam onto each other’s 2G and 3G networks in rural areas
would save significant network costs while enabling greater
network coverage. In fact, competitors have even started sharing the bulk of their wireless access network facilities in nonrural areas; one example is Telstra’s and Hutchison’s shared 3G
network in Australia. Similarly, France has allowed infrastructure sharing among 2G operators in order to reach unserved
rural areas known as zones blanches. Such roaming and infrastructure-sharing arrangements could also apply to new broadband wireless networks.
C HAPTER 4
Fibre backbone networks are scarce in most developing
countries, making broadband deployment more challenging. Fibre backbones can boost the capacity of DSL networks.
Extending fibre closer to rural areas can also facilitate internet
backhaul for wireless broadband technologies. Again, rather
than resorting to a “super” fibre backbone operator, regulators
can promote synergies between different kinds of utilities or
projects that employ internal communications links. Energy
and transport infrastructure projects, such as electrical plants,
highways, railways and pipelines, could be encouraged to
deploy fibre as part of their projects. Telecommunication operators could then access these fibre facilities to augment their
networks.
Regulators could also provide incentives, such as tax
breaks, for 2G mobile operators to build their own backbones
using fibre, instead of the more commonly used microwave
links. The regulatory framework could allow owners of such
communication resources to lease unused capacity to others
for commercial deployment.
In addition, countries are introducing new regulatory
tools to encourage network investment by smaller market players. Ireland, for example, has found that rather than imposing national broadband rollout and coverage obligations on
large-scale operators, it can achieve greater success by allowing
wireless broadband providers to enter small local service areas.
Ireland’s practice of licensing small local service areas--defined
as a 15-kilometre radius around a base station—has led to a
significant rise in new broadband subscribers in non-urban
areas.11
Encouraging competitive market entry is part of a larger
package to promote broadband deployment by a full range of
potential broadband providers. Additional elements of this
larger package are explored below.
4.4.2 The Role of Foreign Ownership
New market entry and subsequent investment, including
in broadband access networks, is likely to be supported if there
are no restrictions on foreign ownership of licensees. This is
particularly true in developing countries, where capital availability may be limited. Foreign ownership brings the possibility
of incremental capital funding, as well as managerial expertise
and international best practices. Increasingly, governments and
57
regulators seek to attract foreign ownership, rather than restrict
it on the premises of national security, cultural protection and
domestic economic development (although these often remain
concerns for policy-makers).
•
Have an efficient administrative process that is transparent
and consistently applied, together with minimal administrative requirements;
•
Ensure that any terms and conditions included in licences
are not financially punitive and allow operators to achieve
sufficient financial return over the life cycle of their
investments;
•
Ensure that licensing fees are commensurate with the
required activities of the licensee – that is, if the licensee is
expected to deploy substantial broadband access infrastructure, fees should be reduced to reflect this high level of
investment, rather than treating operators as “cash cows”
to be milked for government revenues;
•
Require regulators to establish and enforce appropriate
monitoring mechanisms to ensure that licensees meet
their commitments or specific conditions of license agreements;
•
Recognize that re-negotiating licence commitments
increases the risk associated with network investment to
promote broadband access deployment and take-up.
There are persuasive reasons to facilitate and support partial and even full foreign ownership of new entrants in many
circumstances. These reasons include:
•
There is an international trend to ease foreign-ownership
restrictions, partially encouraged by World Trade Organization (“WTO”) agreements, but also driven by increasing
recognition of the substantial benefits that foreign ownership provides;
•
It is possible to address investment requirements through
licence commitments, ensuring that these are not so high
as to lead to excess capacity; and
•
Foreign investors are more likely to see greater investment
risks in situations where they lack managerial and operational control. Greater financial ownership is likely to be
associated with greater managerial control and, by extension, enhanced company performance.
In addition, there are increasing examples of foreign ownership of new entrants that result in substantially improved
telecommunication infrastructure. Foreign ownership will not,
per se, result in greater broadband access network deployment.
But it may support such deployment through greater access to
capital, more managerial experience and, potentially, lower unit
costs.
Besides allowing foreign ownership, governments can also
tap into global capital markets themselves and tap international
lending agencies for funds to improve and upgrade their telecommunication networks. This is much less difficult than ever
before, because it has been proven decisively that operating telecommunication networks – when they are properly deployed
and managed – is a commercially lucrative business.
4.5 Broadband Licensing
Regulators can often affect the rate and prevalence of
broadband network build-outs through their licensing provisions. In general, regulators are experimenting with new
licensing frameworks, often eliminating operator-specific conditions in favour of generally applied regulatory codes. Licensing frameworks can also be employed to provide incentives or
direct mandates for network deployment.
4.5.1 Consistent Licensing and Authorization
Frameworks
The deployment of broadband can be expedited through
relaxing the licensing conditions for large-scale broadband
access providers and by establishing a consistent licensing
framework that is clearly targeted to achieve a set of defined
policy objectives. In addition, regulators are increasingly using
general authorizations in lieu of onerous licensing regulations
to ease market entry.12 Where licences are distributed, it is
important to:
58
While licences, or at least general authorizations, are usually required for large-scale broadband infrastructure operators,
regulators are increasingly lightening such requirements for
operators and service providers in small, rural areas. Facilitating broadband market entry in these areas allows broadband
providers to test their broadband business cases on a small
scale. Some small-scale broadband providers may later decide
to commit to more large-scale deployment. Thus, regulators
can replace licensing requirements for commercial community
broadband providers by a general authorization or registration
framework, just as some countries have already established
“open entry” policies for internet service providers (ISPs).
Where broadband access will be used exclusively for
public services, such as in health facilities or libraries, regulators may question whether licensing should apply at all. It is
also particularly important that licence fees for very small
broadband providers be kept as low as possible, if not eliminated altogether. Licensing obligations that may apply to largescale operators, such as rollout and coverage obligations or
contributions to universal access funds, can be minimized or
eliminated in a regulatory framework targeted to community
broadband providers.
A case can also be made for allowing resale of broadband
services without any licensing requirements in rural areas. For
example, broadband subscribers in a rural area could be allowed
to use their broadband connections to set up a public kiosk and
resell the service. The customers of these kiosk services might
not otherwise be able to afford service at all, at subscription or
monthly rates. In this way, additional economic activity would
be generated through increasing broadband access.
It is important to note that reducing or eliminating licensing requirements is not synonymous with ceasing to regulate
service providers. In some countries, for example, telecommunication licensing is not widely used as a regulatory instrument. Instead, regulatory rules are enacted through universally
applicable regulatory codes, decisions or orders. Even with
C HAPTER 4
Box 4.4: Licensing Incentives for Network Deployment, selected examples
Hungary
In February 1994, Hungary was divided into 54 franchise areas for local telecommunication access. The incumbent and
new entrants were invited to submit bids for each franchise area to act as the monopoly local access service provider until January
2002.
The licence conditions required each operator to achieve annual growth in local access lines of 15.5 per cent, per year. By
January 1997, licensees also had to fulfil 90 per cent of customer demand for new local access lines within six months.
In 1993, teledensity was 14.5 per cent (14.5 per 100 inhabitants). This had increased to 26 per cent by year-end 1996 and to
36.1 per cent by year-end 2002. After 2002, all service providers with significant market power were required to provide local loop
unbundling to other service providers, including new entrants (including a special Reference Interconnection Offer or RIO for
unbundled local loops).
Brazil
Following the deregulation of the telecommunication sector in 1997, operators that met their universal service obligations
were allowed to acquire additional licences, including mobile and long distance service authorizations.
In early 2004, the regulator Anatel certified that Brasil Telecom had met its universal service targets. This allowed the operator to roll out mobile and long-distance services in addition to the local service it was already offering in the southeastern part of
Brazil.
Meanwhile, Brazil’s other two major landline operators, Tele Norte Leste and Spain’s Telefonica Internacional, had already
met their targets and were offering wireless phone and long distance services in other regions by 2004.
open-entry or simple notification policies, local commercial
broadband providers could still be subject to government oversight in areas such as consumer protection and spam. Again,
they could be treated like ISPs, which often come under general business regulation that applies to all commercial entities
— or at least a certain group or “class” of companies.
4.5.2 Using Licensing To Encourage Broadband
Deployment
For example, the right to serve rural areas could be bundled with more lucrative locations in licence tenders.14 In addition, multiple services can be bundled under one licence or
authorization. In 2005, for example, Brazil announced that it
would issue licences enabling operators to offer the “triple play”
of voice, internet and broadcast services. Such an incentive
framework has been used to encourage network deployment
in other countries, as well. For example, in Hungary, the incentive was possible extension of a licence period (See Box 4.4).
Regulators can use licensing frameworks to provide incentives for network deployment by large-scale operators, especially in early stages of market liberalization. This often works
particularly well with respect to the deployment of broadband
access technologies.13
Along with incentives, regulators can also threaten to
revoke licences if commitments are not met. There is, however, a natural tendency on the part of regulators not to revoke
licences, especially if doing so would cause disruption to significant numbers of subscribers.
The intent of this approach should be to encourage operators to deploy networks that may otherwise not be considered
commercially viable or may create less value than other options.
The incentives can either be rewards for meeting licence commitments or (where they can be enforced) financial penalties
for failing to meet agreed commitments.
Such licensing incentives could take several forms:
•
Extension of licence periods;
•
Access to other operators’ infrastructure;
•
Allowing the provision of other, more lucrative services
under the same licence;
•
Access to universal access/service funds;
•
Reduced licence fees;
•
Tax incentives, including reduction of taxes and duties for
both operators and end users; and
•
Financial penalties for failing to meet licence commitments.
C HAPTER 4
Also, in addition to providing tax incentives such as tax
“holidays” or tax concessions to operators, reducing the burden
of taxes and duties on equipment can lower costs for end users,
encouraging wider adoption and usage. A recent study commissioned by the GSM Association indicated that in many developing countries, up to 20 per cent of the total cost of mobile
telephony stems from taxes and duties. Reducing or abolishing
these would allow more people to afford services.
To date, most incentive mechanisms have been applied
only to traditional local telephony (PSTN) services, but there
is no reason why they could not also be applied to encourage
deployment of broadband access networks. For example, new
entrants (or even existing operators) could be offered appropriate incentives to deploy all types of broadband access technologies, especially in rural areas. Given the relative success of such
incentives for PSTN deployment, they could have a positive
impact on broadband deployment, especially if they were available to all industry players.
59
On the other hand, using exclusivity periods as an incentive mechanism poses the danger of “crowding out” potentially
more efficient new entrants and new investment sources. As
long as licensing frameworks are appropriate, market liberalization generally will yield greater benefits than the exclusivity
approach (often coupled with build-out obligations) adopted
in many markets – both developing and developed.
4.5.3 Technology Neutrality
In principle, broadband regulation should be technologically neutral, and licensing and authorization regimes should
reflect this. Increasingly, licensing focuses on generally encouraging the construction of – and investment in – broadband
access networks, rather than defining the specific method of
delivering broadband services. This principle is particularly
relevant to spectrum licensing for broadband wireless services,
but it also can apply to wire-line deployment, giving licensees
the flexibility to use copper, fibre or hybrid networks.
Technology- and service-neutral licences and authorizations also enable broadband providers to offer a full range of
services (including the “triple play” of voice, internet and video)
in rural areas, increasing revenue stream options. In Venezuela,
for example, rural licences allow operators to offer mobile and
multimedia services in addition to fixed access, long-distance
and international services. India and Uganda have allowed
operators to provide both fixed and mobile services under the
same licence, leading to increased competition and subscribers as well as lower prices for consumers. Hong Kong, China
is issuing unified licences for broadband wireless access providers, allowing them to adapt to technological developments.
Today, the licences allow broadband wireless access providers
to offer fixed services, but as BWA technology develops, the
same licences can be used to offer mobile broadband wireless
services, as well.
Pure technology neutrality in licensing may be quite difficult to achieve. For example, some wireless technologies and
services are specifically reliant on, and standardized for, certain
radio frequencies. But regulators are increasingly providing
licensees with the maximum flexibility possible to select which
technologies they wish to adopt, within approved standards
and international frameworks.
Such flexibility may encourage broadband access deployment in marginal areas and developing countries by allowing
licensees to select the delivery technology that most minimizes
costs and accelerates financial return. Operators can customize the components of their network infrastructure to suit
the particular service offerings and technical requirements of
their business plans. This will allow them to leverage whatever
existing economies of scale they have been able to achieve in
other, possibly adjacent, markets. Such flexibility is a hallmark
of India’s unified access service licensing framework, which
gives operators a choice of using either GSM or CDMA within
their assigned spectrum blocks (See Box 4.5).
Some countries specify the use of particular technologies (often made by specific companies) as a tool for industry
development or trade policy. But mandating use of a specific
technology within a defined spectrum block – or for particular
60
kind of service – may not result in the most efficient allocation of spectrum or the most rapid deployment of that service. For example, if the required technology turns out not to
be the optimal one, in terms of cost or supported functionality, market take-up is likely to be disappointing and spectrum
usage will be limited.
This does not imply, however, that technical standards
or international radio frequency allocations are unimportant. Employing a common technical standard provides many
advantages, including:
• Economies of scale (both in terms of network and enduser equipment);
• Commitments and support from large-scale vendors;
• More consistent road maps for product evolution and
development;
• More inter-operability;
• Reduced consumer switching costs; and
• International roaming capabilities.
Standards also provide agreed-upon best practices that
can drive more efficient usage of spectrum and energy/power
– two particularly important elements in resource-scarce areas.
These advantages have often accelerated deployment of telecommunication services above and beyond what could have
been expected if technologies had remained fragmented.
Nevertheless, accepting the benefits of standards does not
mean that regulators should specify which standards should be
used in any specific spectrum allocation or service. Nor does it
necessarily mean that regulators should limit operators’ technology selections to internationally standardized technologies.
The regulatory framework could provide licensees with the
flexibility to select the appropriate technology for their circumstances in order to encourage the deployment of broadband
access infrastructure. There is, of course, the risk that licensees
may choose to deploy non-standard, unique and proprietary
technologies, but given the substantial benefits of internationally recognized technologies, this is unlikely in most cases.
4.6 Alternative Approaches to Broadband Deployment
Regulators should not limit themselves to considering
only current wire-line telephony and mobile network operators
as potential broadband service providers. New market entrants
can be allowed, and even encouraged, to enter the broadband
sector, either by building or converting their existing infrastructures into broadband platforms, or by taking advantage of
unbundling and infrastructure sharing.
4.6.1 Alternative Broadband Platforms
Aside from broadband wireless and fixed-line networks
of major telephony providers, there are other platforms that
can be used to provide broadband voice and data transmission
and internet access. As discussed in Chapters 2 and 3, cable
TV systems, satellites and electrical power grids all offer current and potential alternatives. Cable modem technology has
already demonstrated its role as a broadband access technology
C HAPTER 4
Box 4.5: Unified Licensing Frameworks, selected examples
India
In 2004, The Telecom Regulatory Authority of India (TRAI) established a unified licensing regime for local access networks,
in response to an increasing overlap between GSM mobile services and CDMA “limited mobility” offerings. Now in its first phase,
the Unified Access Service Licensing (UASL) framework covers all basic and cellular services. In the next phase, the government
plans to implement a fully unified licensing regime covering all telecommunication services.
This single-licence approach reduces the financial burden on operators, in terms of licence fees and registration charges,
for offering multiple services. The UASL represents a significant lowering of entry barriers for new and smaller market players,
potentially translating into lower prices for end-users.1
Nigeria
In February 2005, the Nigerian Communications Commission proposed reviewing Nigeria’s telecommunications regulatory
framework, with the goal of establishing a unified licensing regime. Under the new regime, new and existing fixed wireless and
mobile licensees will be allowed to provide both fixed and mobile services, subject to geographical/regional limitations stated in
their licences.
With the removal of fixed-mobile differentiation, licensees will be free to offer voice, data or multimedia services as they
deem fit, once spectrum is allocated.2
1
See ITU Trends 2004, Chapter 5. TRAI press release August 2004. http://www.trai.gov.in/Newpressrelease.pdf
2
The Nigerian Communications Commission, 2005. http://www.ncc.gov.ng/Headlines/REPORT%20ON%20POST%20EXCLUSIVITY.doc
– in fact, it remains a major market leader in North America.
Satellite broadband access remains limited, to date, but it nevertheless is considered a viable option for developing countries.
Broadband over power line (BPL) is often discussed, although
the extent of its viability has yet to be proven in developing
countries, where electrical grid infrastructure may be as scarce
as wire-line networks..
In more developed countries, a primary driver of broadband take-up is competition between telephony (PSTN) operators and cable system operators (for example, in Hong Kong,
China; Korea (Rep.); the United Kingdom and the United
States). In many of these countries, in fact, cable systems were
the first movers in the broadband consumer market, prompting tardy incumbent telephone operators to respond with
xDSL. Furthermore, in the absence of large-scale local loop
unbundling access, the only substantial competitive threat to
telephone operators still comes from operators of alternative
access networks – primarily cable TV systems.
Cable infrastructures are more limited in developing
countries although they have been deployed in a number of
upper middle-income economies (e.g. Hungary), some lower
middle-income countries (e.g. Thailand) and even in some
low-income economies (e.g. India). In these markets, however,
cable is present mostly in urban areas and even there, mostly
in higher-income neighbourhoods. Fundamentally, cable TV is
perceived as a high-income market. It is also a relatively expensive technology to deploy, given the infrastructure requirements. Options for lowering costs (for example, using poles
rather than burying lines) often expose systems to risks from
theft or damage. In some countries, cable systems suffer from
C HAPTER 4
poor QoS and are not robust enough to offer reliable broadband access. But this has not deterred some operators, where
they have infrastructure in place, from offering cable modem
services.
Considering its lower cost structure, satellite broadband
delivery has attracted attention, although market take-up
remains limited. Satellite services are increasingly offered, however, in many developed countries, including Canada, Ireland,
the United Kingdom and United States. Satellite operators
such as Eutelsat, Hughes Network Systems and Shin Satellite
have been aggressively promoting satellite broadband solutions.
Box 4.6 provides examples of satellite broadband deployments
in developing countries.
The primary advantage of satellite service provision is that
it avoids the high cost of terrestrial backbone infrastructure.
Furthermore, the VSAT terminals required to access the satellite broadband service can use battery or solar power, eliminating the need for connection to the power grid. This allows
rapid provision of broadband access to all areas of a country,
assuming that the satellite footprint is sufficient and there is
sufficient transponder capacity.
The key drawback of satellite platforms, however, remains
limited space-segment capacity. New satellites are expensive to
design, build and launch, and transponder lease prices can be
similarly expensive as a result. By contrast, small VSAT dishes
have declined in cost substantially in recent years.
Similarly, although power-line broadband technology promises to support the delivery of telecommunications
services over electrical infrastructure (with only minor modifications), BPL remains largely in the trial stage. The use of
61
Box 4.6: Satellite Broadband in Developing Countries, selected examples
Algeria:
Provision of satellite broadband to enterprises and public organizations in rural areas.
Ethiopia:
Schoolnet is a programme funded by the United Nations Development Programme (UNDP) to provide satellite
broadband to 400 schools.
Thailand: Commercial deployment of satellite broadband services and delivery to schools.
Uganda:
Trial provision of satellite broadband services to number of rural schools.
existing electrical power infrastructure, however, could reduce
the substantial investment requirements to build or upgrade
telecommunication networks. This potential has attracted a lot
of attention as a mechanism for providing an alternative fixed
local loop with broadband capabilities. Of course, as already
mentioned, BPL is intrinsically dependent on existing electrical infrastructure, which is notably absent in large areas of
developing countries.
4.6.2 Wholesale Provision and Unbundling
The deployment of broadband access service has, to a
certain extent, been encouraged in more developed countries through various forms of local loop wholesale products,
including full local loop unbundling and line resale. These
products allow new entrants to access end-users without sinking the investments required to build local loop infrastructure.
The various forms of local loop wholesale access are discussed
in Box 4.7.
Success in providing wholesale local loop products has
varied in different countries, but in Hong Kong, China, the
United States and Germany, unbundled local loops now
account for a significant proportion of direct exchange (or
access) lines. The establishment of appropriate unbundled loop
access prices, at levels that permit incumbents to retain a moderate margin, has encouraged alternative operators to employ
unbundling to provide broadband services to end-users. This
has played a significant role in creating demand for broadband
access.
In developed countries, the local loop has commonly been
constructed before market liberalization. Moreover, the incumbent usually has had a significant period of time as a monopolist to recover the upfront infrastructure costs. So in developed
countries, unbundling has been less a tool to promote further
network build-out or teledensity than a technique to encourage service competition. So developed countries have been
able to price unbundled local loop access based on incremental
costs of offering a wholesale service, rather than historical costs
of building the loop infrastructure. The situation in developing
countries is fundamentally different for the following reasons:
1 Many developing countries are liberalizing primarily to
achieve network expansion and higher teledensity – not to
increase service competition;
2 Liberalization is occurring much earlier in the development chain – without landline networks being fully constructed;
62
3
4
Many incumbent operators may not have had sufficient
time to recover the costs of local loop network deployment that has incurred, and therefore, pricing based on
incremental cost methodologies may not be appropriate;
Use of fully allocated cost approaches may have greater
justification and validity for loop access pricing in developing countries, but these will result in higher wholesale
prices (not least because they may reflect the inefficiencies and historic network structures of the incumbent),
reducing the attractiveness of local loop wholesale access
to alternative operators.
In addition, regulators in developing countries may lack
administrative expertise and resources to implement and regulate a wholesale market structure. Even in developed countries,
regulators have struggled to establish a local loop wholesale
framework that encourages the development of market competition. Many incumbents have successfully impeded wholesale access by setting up technical or limitations to co-location,
slowing down processes for provisioning wholesale loops or
pricing loops at levels that make a mockery of alternative operators’ business cases.
Not surprisingly, few regulatory authorities in developing
countries have opted to build broadband deployment strategies around the provision of wholesale local loop products by
incumbent operators. Those that have opted to implement
local loop unbundling recognize that its success rests on the
ability to enforce the associated requirements. Although the
provision of wholesale local loop products remains very limited
in developing countries, a few developing markets are planning
and implementing local loop access mandates. In September
2005, the members of the West African Telecommunication
Regulators Assembly (WATRA) agreed to a set of regulatory
guidelines that includes support for bit stream access.15
The provision of wholesale local loop products, including
local loop unbundling, will meet with greater success in promoting the deployment of broadband access networks where
certain regulatory and commercial conditions are in place.
These success factors include:
• The existence of an extensive and well-developed incumbent network;
• Clear and complete regulations that spell out all unbundling requirements to ensure that strong operators do not
impede access to their exchanges; and
C HAPTER 4
Box 4.7: Local Loop Wholesale Options
There are three main types of local loop wholesale access:
Local loop unbundling allows access seekers to have management control over the copper pairs connecting a subscriber to
the incumbent’s main distribution frame (MDF). The access seeker can provide both voice and data services on the incumbent’s
network.
Shared access refers to an arrangement where competitive service providers have access to either voice or data transmission
over the incumbent’s network. The access seeker leases part of the copper pair spectrum while the incumbent maintains control
of the copper pair.
Bit stream (or wholesale) access involves the incumbent installing high-speed access links to its customers and opening
these links to competitors. In this case, the access seeker has no management control over the physical line and is not allowed to
add any equipment to the network.
Total Service Resale allows an alternative provider to purchase the network operator’s service at a wholesale discount,
rebrand the service, and resell it to the consumer, allowing the alternative provider to build a customer base and obtain a retail
sales margin over the wholesale rate.
Box 4.8: Local Loop Unbundling in Poland
In February 2005, the Polish Office of Telecommunications and Post Regulation issued a directive requiring the incumbent telecommunication operator, Telekomunikacja Polska (TP), to give other operators access to its local loops. The move was part of a plan
to further liberalize the Polish telecommunication market.
In the first phase of local loop unbundling, TP will provide both full (voice and data) and shared (data) access to competing
operators. There also are plans to extend the local loop unbundling offer to include bit stream access. The regulator is still reviewing
the cost model submitted by TP, but local loop prices are expected to be based partly on benchmarking against other European
providers operating in competitive markets.1
1
•
See http://europa.eu.int/information_society/topics/telecoms/implementation/annual_report/7report/slides281101
Conditions that encourage continuing investment by both
incumbents and new operators in new infrastructure rollouts.
Regulators may decide to end wholesale local loop
requirements once new operators achieve an appropriate level
of commercial scale. Regulators can then place more emphasis
on frameworks that encourage network deployment.
entrants. These bottlenecks often stem from either a lack of
needed infrastructure, or from monopoly ownership of all
infrastructure by a single operator – typically, the incumbent.
For example, in some markets, incumbents have used their
ownership of backbone infrastructure to impose constraints on
the ability of new entrants to compete by:
•
Imposing excessive access/leasing costs – In some markets, for
example, incumbents levy higher leasing charges on competing telecommunication operators than on their corporate customers;
•
Imposing service limitations – Some network operations try
to limit the amount of bandwidth or QoS they will support; and
•
Imposing restrictions on points of interconnection – Some incumbents set technical restrictions or limit access to international switches.
4.7 Competition and Industry Regulation
This section explores how regulators can address the various issues that arise as broadband markets evolve and mature.
As with telephony markets, broadband markets may undergo
transitions from monopoly or near-monopoly structures to
greater competition. Moreover, they evolve in the context of
existing and allied telecommunication markets, such as voice
telephony.
4.7.1 Backbone and International Connectivity
Issues
Multiple network bottlenecks can occur in the provision
of broadband services, both by incumbents and new market
C HAPTER 4
In a number of markets, control of backbone and international connectivity (whether by incumbents or other monopoly operators) has been used to manipulate and constrain the
development of competition. This can lead to artificial shortages of bandwidth and inflated prices, thereby hampering the
provision of robust global telecommunication services. Due
63
to such concerns, the regulator in Singapore, for example, has
moved to open up access to submarine cable landing stations.
There are two main options available to regulators to head off
the adverse consequences of monopoly control of backbone
infrastructure:
1
Impose a tighter regulatory framework on owners of bottleneck facilities to ensure that other operators can access
such backbone infrastructure at an appropriate pricing
point; and/or
2
Encourage existing or new licensees to deploy alternative
infrastructure.
Regulatory intervention is often required where the costs
and timelines associated with duplicating bottleneck facilities
would be so excessive as to be commercially nonviable. Some
countries now permit new entrants to install VSATs with international access. Some also allow local broadband providers to
connect directly with international backbone networks, rather
than terminating traffic through an incumbent’s international
gateway. But in many developing countries the incumbent
still controls access to international network infrastructure
and is able to use this control to impose excessive prices on
other operators, undercutting competition. So regulators are
increasingly stepping in to ensure that new entrants gain fair
and competitive access to existing backbone infrastructure. In
addition, regulators are finding that reliance on international
connectivity can be reduced by developing internet Exchange
Points and local caching, which the government can actively
encourage or establish.16
It is also necessary to encourage the construction of supporting backbone networks, especially if the intent is to deploy
infrastructure into rural areas. However, such requirements
are more likely to be effective if they are not so prohibitive
as to deter market entry. With respect to encouraging licensees to deploy alternative infrastructure and, furthermore, to
deploy infrastructure into areas previously not accessed by telecommunication networks, there are a number of options and
approaches for regulators to consider:
•
Facilitating access to existing telecommunications infrastructure
used for alternative activities—For example, railway signalling
or pipeline monitoring activities involve communications
links, which can be made available to licensed telecommunication operators.
•
Ensuring and facilitating access to government land, including
railways, electrical grids and road networks – Governments
can streamline and standardize the application process for
access to rights of way and ensure just and reasonable fee
structures.
•
Ensuring that telecommunication networks are incorporated into
new infrastructure developments – Governments can include
broadband network conduits in road projects or incorporate cable arrays in new electrical grids (for example,
Chile’s ICT project side-stepped difficulties of geographical isolation and infrastructure shortage in rural areas by
taking advantage of electrical plant).
•
Creating broadband alliances to pool financial and other resources
and to enhance negotiating power with network vendors – The
64
Wireless Broadband Alliance gives operators in the United
States, the United Kingdom and the Asia-Pacific region
benefits of scale in areas such as testing of products and
services, influencing development and adoption of technology standards and negotiating international roaming
agreements.
Experience from some more developed markets (for
example, Australia and the United Kingdom) shows that backbone networks can be more quickly established by leveraging
existing infrastructure. For example, cables can be strung across
electrical pylons that can also act as the location of radio antennas for wireless technologies. Other alternative infrastructures
include main roads, gas and oil pipelines, and water channels
(especially maintained canals).
The key remaining issue, however, is what to do where
such supporting infrastructure does not exist. The lack of basic
support infrastructure (electrical grids, railways and pipelines)
in rural areas is often heightened by large distances or rugged
terrain. In such circumstances, the regulatory framework can
promote the use of wireless technologies – satellite trunking
– especially where traffic is not expected to be substantial. This
can be achieved through the expeditious allocation of unused
or little-used spectrum and by lowering licence fees for trunking in remote areas.
Developing countries may also explore the pooling of
resources to collectively launch satellites that provide broader
regional service, or to back a commercial satellite operator.
Shin Satellite of Thailand, for example, has just launched
its Ipstar satellite, which has a wide footprint across the AsiaPacific region. Although it is a commercial operator, Shin has
extended broadband access to rural areas of Thailand, Laos,
Cambodia and Myanmar.
4.7.2 Funding Broadband Deployment
The need to deploy infrastructure into more marginal
geographies is based on the recognition that without such
access, the Digital Divide will continue and grow – both
between developing and developed countries and between
urban and rural areas. Given the benefits of broadband deployment, including cheap voice communication, regulators are
playing a critical role in seeking to reduce this divide through
the promotion of broadband access deployment.
The first step regulators are taking is to develop regulatory regimes that are conducive to investment and supportive
of commercially viable broadband access network deployment.
In other words, they are specifically addressing the market efficiency gap in broadband provision. The need to first address
the market efficiency gap in promoting ICT development
was explored fully in the 2003 edition of this report. Efforts
to close the market efficiency gap in basic telecommunications
can also apply to the broadband service market. Regulators
should remove unnecessary regulatory burdens and encourage
market demand for broadband deployment, allowing market
forces to promote broadband access wherever possible. It also
can include facilitating market entry by small enterprises and
micro-entrepreneurs, as well NGOs, libraries and local governments. In addition, regulators can encourage large-scale
C HAPTER 4
Box 4.9: Removing Network Bottlenecks in India
The Indian telecommunication regulator, TRAI, is planning to introduce more competition into the market for international
private leased circuits (IPLCs), which are currently “bottleneck” facilities. This will have implications for international telecom
services like International long distance (ILD), as well as internet access and broadband data offerings.
There is a significant lack of competition in the IPLC market at present, forcing smaller operators out of the market for several
services relying on leased line capacity. The reasons for lack of competition include:
• A limited number of landing stations, which are owned and controlled by a small number of operators;
• The ability of IPLC providers – who are also Internet Service Providers (ISP) – to charge other ISPs prices that are
significantly above costs, harming competition the ISP market;
• The ability of IPLC providers also to charge substantial prices to ILD resale operators, with whom they also compete in
the ILD market.
In response to this, TRAI has undertaken, or is planning, the following initiatives:
• Setting a tariff ceiling for various transmission capacities, based on each incumbent’s costs, and removing differences in
IPLC prices based on usage volumes by ILD operators and ISPs.
• Reviewing the need to permit IPLC resale, which had been banned because it was believed that resale would retard
construction of ILD infrastructure in the country.
• Recognizing the need to facilitate access to cable landing stations by new service providers, as well as by new international
cable carriers.
• Planning to facilitate mutual sharing of landing station infrastructure international cable capacities among the carriers.1
1
TRAI, 2004, Fixation of Ceiling Tariff for International Leased Line Circuit (Half Circuit). http://www.trai.gov.in/consultation%20paper-30th%20april%202004.pdf
Box 4.10: Using the Indian Rail and Gas Facilities for Backbone Connectivity
Railtel Corporation of India was set up in 2000 to exploit communication assets lying idle along India’s rail network. Since
then, the company has laid 25,000 km of fibre optic cables along rail lines. Railtel provides leased lines to telecommunication
service providers, along with other infrastructure like tower space and co-location services. In addition, it is also an ISP, operating
a network of internet kiosks set up at railway stations.
The company plans to open an additional 300 cybercafes at railway stations, providing services such as VoIP and video conferencing for local people with no other access to computer equipment or broadband service. Owing to India’s extensive railway
network, this allows broadband access to be rapidly extended into many marginal areas.1
Similarly, Gailtel, the telecommunication services arm of the largest gas transmission company in India, operates as an integrated telecommunication infrastructure provider. The company started leasing bandwidth to telecommunication operators like
Bharti and Tata in mid-2001. It also operates as an ISP, serving corporate and residential customers.
The company has laid an optic fibre/co-axial (OFC) hybrid network along about 8,000 km of natural gas and LPG pipelines,
and it plans to extend the network to 18,000 km around the country. The network currently serves 73 cities across eight states. Due
to infrastructure cost savings derived from overlaying the OFC network on existing pipelines, Gailtel is able to offer broadband
services to its customers at substantially lower costs, compared with its competitors.2
1
http://www.railtelindia.com/
2
http://www.convergenceplus.com/apr03%20expert%20view%2002.html
operators to deploy broadband networks in areas viewed as not
commercially viable in return for access to potentially more
value-creating business activities or other incentives.
It may also be necessary, however, for regulators to establish mechanisms to fund network deployments, especially
if there is evidence that regulatory incentives and lower-cost
C HAPTER 4
network alternatives will not be enough in certain areas. This
shortfall between market-based measures and universal access
is termed the true access gap. In any given area, this true access
gap (which represents, in effect, the failure of the market to
deliver needed services) can only be determined after all
attempts to address the market efficiency gap have fallen short.
65
Box 4.11: GSR 2003 Universal Access Best Practice Guidelines
During the 2203 Global Symposium for Regulators, delegates adopted a broad manifesto for enabling universal access to
telecommunication networks and services, including broadband ones. The best practice guidelines are as follows:
“We, the regulators participating in the 2003 Global Symposium for Regulators, have identified and propose the following
best practice guidelines to achieving universal access to information and communication technology (ICT) services.
A An enabling regulatory environment: the role of governments and regulators
The success of any universal access/service policy is dependent upon political support at the highest level that recognizes the
role of ICTs as a tool for development.
1) It is essential that Regulators exist or be established where they do not yet exist, and that their key role in implementing
universal access policies and promoting competition be recognized and reinforced.
2) A series of policy and regulatory reform measures can be taken to achieve universal access to ICTs. These include:
a) Formulating a national policy that identifies appropriate and realistic universal access/service objectives that take into
account the differences between universal access-public access to ICTs-and universal service-household or private access
to ICTs.
b) Including all citizens, regardless of gender, ethnicity, socio-economic level or geographic location, in national universal
access/service objectives.
c) Reviewing universal access/service policies, regulations and practices periodically to adapt to the evolving nature of ICT
services and the needs of end users.
d) Conducting periodic public consultations to the extent possible with stakeholders to identify their needs and modify
accordingly universal access policies, regulation and practices.
e) Designing universal access policies, regulations and practices in order to create incentives for the private sector to extend
universal access to communications services.
f) Establishing a fair and transparent telecommunication regulatory framework that promotes universal access to ICTs.
g) Adopting technologically neutral licensing practices enabling service providers to use the most cost-effective technology
to provide services for end users.
h) Adopting a framework of interconnection rates linked to costs.
i) Reducing regulatory burdens to lower the costs of providing services to end users.
j) Developing an effective regulatory body responsible for implementing policies directed towards assuring the best quality
reliable services at the most affordable prices that meet the needs of consumers-existing and future.
k) Promoting competition in the provision of a full range of ICT services to increase access, affordability, availability and
use of ICTs.
3) Countries can use regulatory reform as the first step in achieving universal access, recognizing that further steps may be
necessary to achieve ubiquitous access to ICTs, e.g., in rural areas or to users with special needs.
4) Appropriate licensing schemes for rural service providers could be granted to meet the needs of un-served and under-served
areas.
B Access to information and communication infrastructures
1) The lessons learned from the initial experiences developing countries have achieved with mobile cellular services can be
applied to a broader range of ICT services to foster universal access. These lessons include providing services in a competitive framework, using new technologies that offer both innovative services and affordable pricing options (e.g., pay as you
go options such as pre paid cards) to a wide range of end users.
2) Other measures to promote affordable ICT equipment could include national manufacturing of ICT equipment, reduced
customs tariffs and duties, and end-user loans to foster affordability of ICT equipment.
3) A full range of public access options can be developed, including the creation of public telecentres.
4) Local input (including the content useful for local populations) into projects increases their long-term financial sustainability.
5) Educating local people on the benefits of ICTs and their use increases their long-term financial sustainability.
66
C HAPTER 4
Box 4.11: GSR 2003 Universal Access Best Practice Guidelines
C Guidelines in regard to finance and management of universal access policy
1) Universal service funds can be viewed as an option that complements regulatory reform and developed as a mechanism
within a broader market-oriented approach to achieving universal access.
2) Universal service funds can be financed by a broad range of market players, managed by neutral bodies such as regulators,
and be used to kick-start public access projects that meet the needs of the local community.
3) Governments may consider a full range of other financing mechanisms, including tax incentives for ICT providers and end
users.
4) Competitive minimum subsidy auctions could be used, as an option, to reduce the amount of financing necessary for public
access projects financed by a universal service fund.
5) Public access projects can be designed to achieve long-term financial self-sustainability, especially where consideration is
given to innovative low-cost technologies
See: http://www.itu.int/ITU-D/treg
When all inefficiencies have been leached out of the market
through sector reforms, and the true access gap remains – only
then may government intervention be necessary. This intervention can take the form of targeted and limited “smart” subsidies
to spur the deployment of broadband access networks to areas
and populations that would otherwise not be reached.
Regulators have several ways to address the true access
gap, including:
• Licensing special rural operators to deploy broadband access networks in defined geographies – Licensees can be selected based
on bidding for the minimal subsidy required to achieve
specified targets. By licensing such rural operators within
specific areas, regulators can “leapfrog” the gradual diffusion of new technologies from urban to rural geographies;
• Providing funding for local community initiatives to provide
broadband access. Many universal access fund programmes
take a top-down approach, directing fund administrators
(often regulators) to identify the communities for which
targeted subsidies will be made available. But a bottom-up
approach could also be used, allowing communities themselves to apply for funds to deploy their own broadband
access networks. This would help to ensure local community involvement in, and demand for, broadband access.
• Giving Direct and indirect financial support in return for the
deployment of broadband access networks. Governments could
provide tax exemptions to operators that roll out telecommunication infrastructure in rural areas. Where this is
insufficient to attract commercial operators, governments
could offer full or partial subsidies. Alternatively, the government could provide preferential loans to operators for
building broadband access networks.
• Requiring operators to deploy broadband access networks. Again,
broadband deployment mandates could be tied to funded
mechanisms, drawn from government revenues or contributions made by all operators. Contributions could be
either as a flat-rated or set as a percentage of revenues.
The operator(s) responsible for the provision of universal
broadband access in any given area would receive financial incentives or payments for each new broadband access
C HAPTER 4
line installed. But it could also be given the freedom to
determine in which specific locations it would deploy
such infrastructure. Governments may prefer to provide
funds only where the costs of providing broadband service would exceed revenues from that service.
Any government involvement in financially supporting
the deployment of broadband access networks has potential
drawbacks. For one thing, regulators should ensure that the
allocated funds are used for the expressed purpose of deploying infrastructure in marginal areas. This requires the regulator to have the institutional capability to manage and oversee
the allocation process and to provide maximum transparency.
There are examples of such regulatory frameworks, including
Korea (Rep.) and a number of South American countries (See
Box 4.12).
Deployment of broadband access networks by large-scale
network operators – even those employing wireless architectures – may only be commercially viable in many areas of
developing countries (and even a number of areas in developed
countries) if an incentive framework is incorporated into the
licensing approach, coupled with government funding and
initiatives that generate customer demand. One approach is to
use regulatory incentives to support and promote broadband
access network deployment by large-scale operators. Another
approach is to encourage small-scale players to serve local
communities by removing regulatory burdens that often apply
to large-scale operators, and allowing small players to test their
business cases for broadband access and build demand. The
key issue is deciding which approach has the best fit with the
underlying regulatory and institutional capabilities, and which
approach minimizes the extent of government involvement
while maximizing the commitments of private investors.
4.7.3 Enforcing Broadband Market Competition
To ensure a reasonable level of competition in an emerging broadband market, it is necessary to establish a regulatory
framework that prevents anti-competitive behaviour by operators. Regulators need to monitor dominant operators, which
can be defined as those that are capable of acting unilaterally in
67
Box 4.12: Korea’s KII Project
The Korea Information Infrastructure (KII) Project was established in Korea (Rep.) in early 1995 as an avenue to promote
nationwide broadband usage. The ultimate aim of the project was to provide broadband networks to 13.5 million subscribers with
the average transmission rate of 20 Mbit/s by 2005.
To achieve this, the government supplied public funding to facilities providers to partially ease the burden of investment in
access networks. Greater funding was reserved for remote areas. In addition, some enterprises were offered a tax exemption on
deployment of broadband infrastructure.
The government provided loans, at preferential rates, worth USD 77 million to facilities service providers in 1999. It committed an additional USD 77 million in 2000 for the purpose of deploying infrastructure in less densely populated areas. Subsequently, public funding was extended to cover infrastructure build-outs in rural areas, with additional investments amounting to
approximately USD 900 million.1
1
Ministry of Information and Communication, Korea http://www.mic.go.kr
Box 4.13: Use of Universal Access Funds, selected examples
In 1995, the Chilean government established a Universal Access Fund to promote installation of public telephone systems
in isolated rural areas. The USF employed competitive bidding to determine which provider should receive the subsidy in any
area. The fund managed to secure approximately USD 60 million in investments, of which about 85 per cent came from private
companies. With support from the fund, public telephones were provided to about 6,000 rural localities, reaching 2.2 million
inhabitants within seven years. Reliance on market forces and minimal regulation, among other things, has been credited for the
success of the scheme.
In Peru, universal access funds for the Fondo de Inversión en Telecommunicaciones (FITEL) are collected through a 1 per
cent tax on the gross revenues of all public telecommunication companies. The funds are then allocated through public competitive bidding to operators deploying infrastructure in locations of priority social interest.
the market (such as on pricing or provisioning terms), without
regard to discipline from competitors, buyers or sellers. Such
regulation need not be onerous or excessively restrictive, it just
needs to be effective in prohibiting and punishing conduct that
has the potential of preventing or constraining the development of market competition.
Examples of potential business conduct by dominant
operators that should be prohibited include:
•
Predatory pricing (providing services at less than cost);
•
Mandatory product bundling (requiring end users to take
products in which the operator is not dominant in order
to access products in which the operator is dominant);
•
Price discrimination (applying different prices and terms
and conditions to favour or disadvantage particular customers); and
•
Cross-subsidization (using profits generated in one service market, in which an operator is dominant, to subsidize
its operations in a competitive market).
In response, obligations typically applied to dominant
operators include:
•
68
Requirements to provide interconnection to competing
operators on appropriate terms and conditions; and
•
Obligations to ensure that tariff structures comply with
regulatory requirements, including appropriate price controls.
All of these issues are relevant to the broadband access
market. A dominant operator – usually the incumbent telephone operator – can distort the development of the broadband access market by undertaking behaviour to stifle or
disrupt competition. Common techniques are predatory
pricing and cross-subsidization, both of which can undercut
smaller broadband access providers’ revenues and drive them
from the market.
In general, developing broadband markets should be subject to ex-ante rather than ex-post regulation, given the potential
for anti-competitive conduct by dominant operators. Ex-post
regulation tends to be work better when there is no dominant
carrier, and where there is sufficient evidence that competitive
market forces will function properly without extensive ex-ante
rules. Given the current structure of most telecommunication
markets – including broadband access markets – such competitive circumstances do not commonly exist.
Even within a regulatory framework that applies more
rigorous scrutiny to dominant operators, it remains important to avoid excessive regulatory burdens. Regulations can
be carefully tailored and minimized to address key concerns:
C HAPTER 4
(a) preventing anti-competitive business conduct, (b) ensuring compliance with licence commitments and (c) protecting
end users. Strict price regulation practices, for example, can be
replaced by a general price cap regime, with the goal of ending
price regulation when the market is genuinely competitive. At
the same time, market development of new services also may
hasten the end of price regulation. The rise of new offerings
for VoIP, which is usually not price-regulated, may force dominant operators to lower prices to retain market share.
Section 4.6 of this chapter has already detailed the role
of competition between cable TV and telecommunications
companies in driving competition and take-up in the broadband access market. Therefore, cross-ownership of telecommunications and cable network companies can serve as an
impediment to broadband development. Integrated providers
that own both telephone and cable TV networks are unlikely
to deploy both DSL and cable modem services, because the
two would cannibalize each other. While good for the operator, deploying only one of the technologies would bring higher
prices for consumers.
4.7.4 Other Regulatory Concerns
This chapter has identified the need to regard broadband
access as an information delivery platform. As broadband takeup and usage widens, regulation of broadband services and
applications will become an increasing concern for regulators.
In view of the potential and growing demand for IP telephony,
regulators in several countries have begun developing a regulatory framework to address issues associated with VoIP. These
key concerns are examined in detail in Chapter 6.
In addition, regulators will also need to consider regulatory measures pertinent to any content delivery platform as
broadband deployment takes off. Given the superior functionalities of broadband access, the need for regulating broadband
may be even more pressing than for other telecommunication
services. Content-related concerns include:
• Having an appropriate framework to effectively guard
against intellectual property infringement;
• Setting content regulation guidelines to protect consumers, especially minors; and
• Setting up appropriate and suitably non-invasive regulatory mechanisms to ensure that national security is not
compromised.
4.8 Increasing Broadband Awareness
So far, this chapter has focused on the role of regulators
in stimulating the supply side of the broadband access market
– that is, encouraging network build-out. There are, however,
significant opportunities for regulators to promote demand for
broadband through the expansion and improvement of applications and services available to end users.
4.8.1 The Government’s Role
Given scarce capital resources in many developing countries, regulators (and appropriate ministries) often must demon-
C HAPTER 4
strate the relevance and necessity of promoting the deployment
of broadband access networks, especially if funding is required
in some form. This requires a clear demonstration of the benefits of broadband access and the preparation of a holistic ICT
strategy. Some countries have used e-government projects to
initiate and support ICT projects, including broadband development. Such projects have several objectives, including:
• To enhance the efficiency of government by converting
paper and manual transactions to online ones, simplifying
and speeding up government processes;
• To improve government linkages with the population and
give citizens more ready access to government services,
officials, and information;
• To expand the reach, awareness and understanding of
broadband access among the population – including the
government’s own employees (the public sector is often
the largest employer in developing countries); and
• To provide a framework to attract local and foreign investments.
The ITU is working on a Global E-Government Project
to enhance government services through the use of secure and
trusted internet infrastructures and applications in selected
developing countries.17 The Vietnamese government, meanwhile, is cooperating with the World Bank to implement an
e-government plan, with the idea of promoting sustained takeup of broadband.
Within such projects, the provision of broadband access
networks is just one component of a broader strategy to
improve ICT positioning and provide a positive impact across
the economy.
Telecommunication regulators are in a good position to
initiate greater collaboration and cooperation with other government agencies to promote broadband take-up. In addition to e-government projects, which invite participation by
multiple agencies, inter-governmental working groups can be
established to facilitate the development of broadband infrastructure, particularly in commercially less viable areas where
resources can be shared. For instance, the Singapore ONE
initiative, which was officially launched in 1997, was designed
to be a collaborative effort between government and industry to implement a nationwide broadband network. It was
jointly supported and driven by multiple government agencies,
including the telecommunications regulator, the Economic
Development Board, the Media Development Authority and
a research agency.
Regulators and government agencies can take the lead by
actually building a broadband backbone network. The success of the Singapore ONE initiative was largely driven by the
government’s construction and operation of a core Asynchronous Transfer Mode (ATM) backbone network, which enabled
broadband access to be provided extensively to public libraries, schools and training centres across the island nation. This
served to raise awareness of broadband and drove broadband
take-up across various communities.
In order to really act as advocates for broadband access,
however, regulators must be able to clearly and credibly dem-
69
Box 4.14: Limiting Cross-Ownership in the EU
In 1999, the European Union issued a directive requiring the separation of telecommunications and cable TV operations
into distinct legal entities. It was believed that cross-ownership of telecommunications and cable operations would prevent cable
companies from providing low-priced voice telephony services in competition with telephone companies. In addition, without
cable competition in the internet access market, there would be little incentive for telephone companies to upgrade their existing
networks to full-scale broadband capability via xDSL.
Hence, under the directive, dominant public telecoms network operators were required to run their cable operations as separate legal entities. It was thought that this would prevent emergence of new, anti-competitive bottlenecks and would encourage
competition and innovation in both the telephony and cable TV markets.
Box 4.15: E-Government in Vietnam
The Ministry of Posts and Telematics (MPT) in Vietnam has recently drafted a National ICT Master Plan for 2006-2010 and
an E-Government Master Plan covering the same period. These initiatives will be implemented in cooperation with the World
Bank. The ICT Master Plan aims to achieve:
(1) A wide diffusion of ICTs throughout Vietnam’s economy and society, making it a larger contributor to GDP;
(2) Establishment of a nationwide information and communications network; and
(3) Comprehensive ICT skills development.
As part of the process to improve e-readiness in Vietnam, the ICT project will address the following issues:
• Strengthening the technical and managerial capacity of the MPT in implementing the ICT initiatives;
• Facilitate increased access to telecommunications, in the context of a gradual move towards a more competitive market
environment and private-sector participation;
• Promote greater awareness of ICTs and e-applications in the business community, with the view to encourage businesses to
adopt e-commerce;
• Support enhanced government online presence and content at the national and municipal levels, through interactive and
dynamic portals;
• Roll out e-government services to businesses in areas such as e-procurement and business or land registration; and
• Support extensive training and awareness-raising efforts to encourage diffusion of ICTs in the private sector.
onstrate its advantages. Indeed, the regulator can even take
the lead in coordinating the provision of broadband access to
various government ministries and departments. This is the
case in Singapore, where the regulator, the Info-Communications Development Authority (IDA) coordinates all ICT
procurement for the government. As a further example, the
Turkish Ministry of Health is partnering with ITU to develop
“e-healthcare” systems. Specific initiatives include giving healthcare providers and citizens access to health-related information
via broadband access, and development of a Primary Healthcare Information Systems and Electronic Health Records initiative.
Fundamentally, in order to be able to increase awareness
of broadband access networks and to successfully position
broadband at the forefront of a country’s telecommunication
strategy, regulators need to be able to articulate the potential
benefits, both direct and indirect, within a cost-justified framework. Furthermore, the rationale for broadband is likely to be
stronger when contained within a broader and holistic ICT
strategy (including policies to improve the rate of PC penetration). This is the challenge for regulators, especially when there
are so many other demands on a country’s telecommunication
infrastructure and scarce resources.
Providing content that is relevant to local communities,
particularly in local languages, can also increase the relevance
and potential impact of broadband access networks. Morocco,
for example, is in the early stages of developing Arabic content
for both mobile and internet services. In Laos, a Lao-language
version of Linux-based graphical desktop, along with Lao-language office tools, has been developed and provided to villages
as part of a plan to promote WLANs.
Promoting broadband access requires more than just
access to broadband networks. People in developing countries
also require access to personal computers or other end-user
terminals – and this inevitably raises cost and affordability
issues. Put simply, the question is whether there are low-cost
terminals that are affordable to users in developing countries.
What role can regulators play in improving the penetration of
PCs and portable devices in the developing world? Without a
70
4.8.2 Promoting Broadband Equipment Take-Up
C HAPTER 4
Box 4.16: The Rural-Enlaces Project in Chile
The Rural-Enlaces project is part of an ICT policy to improve education in 3,600 rural schools around Chile, reaching an
estimated 130,000 students. The project involves providing learning aids, such as computers, as well as access to broadband telecommunications.
The proponents of Rural-Enlaces are convinced that technology should be seen as a means to support existing pedagogical
approaches to rural education, rather than as a cultural invasion of the dominant group. Hence, the project aims to involve various
segments of local communities in the implementation process.
For instance, the project is positioned as a professional development opportunity for local teachers. Apart from technological
knowledge transfer, teachers benefit from being directly involved in the selection of educational software and content. In addition,
they are consulted on the design of learning practices that are most appropriate and relevant to everyday reality in their communities. Meanwhile, parents are also involved in helping out with various aspects of project implementation at the local schools.
Box 4.17: Installing Internet Centres in Southern Brazil
Sud Mennucci is a town with 7,500 inhabitants, located 700 km southeast of Sao Paulo. It will soon benefit from a plan to
install two internet access centres, with 10 computers each, to provide internet access to low-income inhabitants.
Three solutions have been proposed to realize this plan:
(1) Have the federal government donate the computers for the centre;
(2) Incorporate the plan into the state government’s Acessa Sao Paulo programme, which aims to provide free internet access to
dozens of municipalities in Sao Paulo state, or
(3) Engage private companies to set up the centre and provide training in the use of technology.1
1
LatinCom, 2005.
solution to the problem of low PC penetration rates, the provision of broadband access networks will have little impact.
In response to these challenges, regulators also have many
options. They can:
One relevant option for most developing countries is to
concentrate PC usage at single locations – whether these are
cybercafés, community centres, schools or government offices.
Often, cybercafés are the primary mode of accessing the internet in many developing countries (See Box 4.17).
•
Seek to maximize investment flows by liberalizing markets and
permitting foreign ownership. This includes allowing broadband providers to offer a full range of services and applications, such as the “triple play” of voice, internet access and
video/multimedia programming.
•
Encourage the deployment of wireless broadband access networks
by freeing up the requisite spectrum. This strategy can be augmented by a technology neutral approach to spectrum
assignments.
•
Create a regulatory framework that encourages a full range of
potential broadband providers. Moving beyond large-scale
national network operators, regulators can empower universities and government offices, local communities and
smaller entrepreneurs to deploy broadband access networks. This may include tailoring regulatory frameworks
to each group of potential broadband providers.
There are, however, many alternative options for encouraging private PC take-up, some of which are detailed in Box 4.18.
4.9 Conclusion
Although basic telephone (PSTN) teledensity is extremely
low in many countries, the rapid development of technologies
– especially wireless technologies – means that policy-makers
and regulators should not lose the opportunity to plan beyond
increasing basic teledensity. They should plan for broadband
access, as well.
Regulators clearly face numerous challenges in the broadband context. In particular, they face a perceived lack of local
demand and available revenue streams for broadband in many
countries. This could prevent the commercial deployment of
broadband access networks in many areas, especially rural areas
– at least by large-scale network operators. Furthermore, continued low PC penetration rates in many developing countries
could effectively negate any potential positive impacts that may
arise from broadband network deployment.
C HAPTER 4
–
A regulatory framework tailored to small broadband
providers will enable and encourage local community providers to harness the potential of broadband
technologies and enable greater broadband access in
rural areas;
–
Competitive large-scale operators can be encouraged to extend their networks to rural areas through
71
Box 4.18: Encouraging PC penetration, selected examples
In India, the government has launched the “Indian PC Programme,” which aims to improve PC penetration from the current
14 per 1,000 to 65 per 1,000 by 2008. Some of the initiatives include: (a) launching INR 9,999 (USD 230) “no compromise” PCs,
subsidized by software vendors and chipmakers; (b) encouraging all incumbent operators to move towards a subscription model for
offering broadband services and PCs as a package; (3) setting up loan schemes, employee provident funds and other saving funds
to encourage PC adoption among government employees; and (4) amending the Income Tax Act to allow deduction of home PC
purchases. Also, cybercafe kiosks have been set up along railway tracks around the country to provide computer access to rural
villages. As a result, rural communities now have access to e-Government, tele-education and telemedicine services.
In Mexico, public internet kiosks were set up in community plazas under the e-Mexico project, which benefited 3,200
municipalities around the country. Each community plaza has an average of 10 computers and has internet access via satellite
technologies.
In Sri Lanka, in cooperation with the World Bank, the Sri Lankan government is planning to set up tele-centres in rural areas
around the country to improve community access to ICTs. The target groups include farmers, students and SMEs.
In Thailand, a “Computer ICT Programme” was launched in 2003 to provide low-cost computers. In addition, there is plan
to establish a nationwide network of 751 tele-centres, located at various post offices throughout the country.
In Tunisia, the World Bank is working with the Tunisian government to subsidize 6,000 “Publinets” and 10,000 PCs under
the “PC Familial” programme. There is also an investment plan to increase the number of PCs in schools.
In Uganda, Uconnect, an NGO based in Uganda, imports used computers from Europe and the United States, revamps
them and supplies them to schools and organizations. About 100 mostly rural-based schools have benefited from this project.
•
•
•
72
infrastructure-sharing arrangements that guarantee
open access to all competitive operators;
–
Competitive large-scale operators can be given
incentives to deploy networks in return for appropriate rewards;
–
Regulators could seek to encourage the deployment
of broadband access networks by providing direct,
targeted subsidies from universal access funds or
indirect financial benefits (such as tax exemptions)
to a full range of broadband providers;
Create an asymmetric regulatory regime to prevent the dominant operator (often the incumbent) from constraining
the development of competition in the broadband access
market;
Work with other government agencies or ministries to develop initiatives, such as e-government programmes, that generate
demand for broadband services;
Encourage the build-out of fibre backbone networks to boost
the capability of both wire-line and wireless broadband
technologies. These steps include forging synergies with
transport and energy infrastructure projects and providing
incentives for 2G mobile operators to replace their microwave links with fibre networks. It also means making it
possible for all owners of such communications resources
to lease unused capacity to others for commercial deployment.
•
Link broadband access development strategies to efforts to support and promote PC take-up. Build government-sponsored
PC kiosks and other access terminals, especially in areas
where broadband networks are to be deployed.
Promoting broadband access in developing countries
requires a new vision of reduced regulatory burdens, innovative incentives, and coordinated efforts by all links in the
broadband value chain. It also requires concerted political will
to achieve. It is an end to the “business as usual” approach,
and governments should treat ICTs and broadband networks
as tools for development. There are many challenges in the
road ahead. Strategic and creative thinking to overcome these
challenges is required, but with concerted efforts on all fronts,
regulators can enable developing countries to join the broadband world.
C HAPTER 4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Large-scale operators include main national operators, new entrants and mobile operators.
Marginal areas include remote, rural and under served areas.
While some private companies have also deployed private networks, these remain largely in urban areas.
By year-end 2004, some 112 countries had at least partially privatized their State-owned incumbent operator. Source: ITU World Telecommunication Regulatory
Database.
The African Virtual University was established in 1996 as part of a World Bank project. It was founded as a technology-based distance education network to bridge
the digital divide in Africa. http://www.avu.org/
Universitas 21 is an international network of leading research-intensive universities, with the objective of facilitating collaboration and cooperation between member
countries, as well as to create entrepreneurial opportunities for them. Established in 1997, it now has 17 member universities in 9 countries. http://www.universitas21.
com/
TRAI, 2004, Recommendations on Internet and Broadband. http://www.trai.gov.in/Recommendations on Internet and Broadband 2004-04-29 FINAL.pdf
Yun et al., 2002, The Growth of Broadband Internet Connections in South Korea: contributing factors. http://www.ciaonet.org/wps/yuk01/
TRAI, 2004, Recommendations on Internet and Broadband. http://www.trai.gov.in/Recommendations on Internet and Broadband 2004-04-29 FINAL.pdf
Contribution of Ireland to GSR Consultation to identify best practice guidelines for spectrum management to promote broadband access (http://www.itu.int/ITU-D/
treg/Events/Seminars/2005/GSR05/consultation.html).
The trend away from onerous licensing practices toward more flexible market-entry authorization approaches, including technology and service-neutral licensing and
authorizations, was explored fully in the 2004/05 edition of Trends in Telecommunication Reform: Licensing in an Era of Convergence (http://www.itu.int/publications/).
See ITU Trends in Telecommunication Reform 2004: Licensing in an era of convergence.
For a detailed discussion on this issue, See ITU Trends in Telecommunication Reform 2003: promoting universal access to ICTs.
The full set of regulatory guidelines agreed by WATRA are available at http://www.itu.int/ITU-D/treg/Events/Seminars/ITU-EC-Project/Ghana/modules/CompilGuidelines_final_E.pdf
The role of regulators in promoting IXPs is explored in the ITU-IDRC report Via Africa: Creating local and regional IXPs to save money and bandwidth, available
at http://www.itu.int/ITU-D/treg/
For more information on this project see: http://web.itu.int/ITU-D/e-strategy/projects/E-Government/Executive-summary11.pdf
C HAPTER 4
73
5 BROADBAND SPECTRUM MANAGEMENT
Author: John Muleta, Venable LLP
This chapter discusses the challenges spectrum regulators face in allocating and assigning spectrum for broadband
wireless access (BWA) and other wireless services. The main
challenge is to provide for flexible, market-oriented spectrum
licence rights, which can create a positive investment climate
for BWA services. At the same time, regulators want to discourage uneconomic hoarding and speculation in spectrum, which
could delay the rollout of services to consumers.
Flexible spectrum rights should be granted as long as
the spectrum licensees meet two absolute preconditions critical to the development of communications markets. First, the
licensee (or, in the case of unlicensed spectrum, the service
provider) must increase competition, benefiting consumers.
Second, licensees and service providers should experience the
opportunity cost1 of using their spectrum allocations. This is
the best way to ensure effective and efficient use of the spectrum.
This chapter aims to explain these concepts and stimulate thinking about how to make effective and pragmatic spectrum management decisions, without falling into the kinds of
dogmatic approaches that have often characterized spectrum
management discussions.2 It is important to understand the
primary goals of spectrum regulation, as well as the fundamental economics of wireless access systems. It will also be useful
to review some of the technological advances that are making
new spectrum resources available for BWA services and applications. The traditional regulatory spectrum management models
can then be examined to determine whether they adequately
address the challenges presented by these new technologies.
Then, best practices can be identified to improve spectrum
management in ways that encourage the rapid deployment of
BWA systems.
below 6 GHz, where the physical characteristics of the spectrum are more conducive to consumer applications.
In this new era, the power of the Information Age to
affect our lives will be exponentially multiplied by the freedom
brought about by BWA networks. Wireless broadband technologies will fuel the engines of the global economy by enabling
consumers to:
• Freely access the internet from the farm, the city, kiosks,
cybercafés, coffee shops, on moving trains, and in their
own communities and backyards, in developed and developing countries alike;
• Connect to the internet seamlessly, using a single device
to make phone calls, access information and government
services, vote, pay taxes and bills, and enjoy entertainment; and
• Live in enlightened communities that are connected to
broadband, spectrum-based services that offer access to
resources and opportunities never before available.
BWA networks may have even more impact on consumers when they are combined with other broadband platforms.
Ultimately, the broadband world will feature BWA networks,
with their mobility and portability, at the core of a variety of
useful hybrid broadband architectures that will provide a rich,
multimedia consumer experience at virtually any time or place.
5.1 Introduction
Of course, the potential for BWA services to improve
people’s lives ultimately depends on the amount of spectrum
regulators make available. But for the first time, it appears that
technological advances can increase spectrum capabilities and
resources, allowing licensees to do more with the same spectrum and enabling entirely new spectrum uses. Regulators
must now consider whether traditional approaches to spectrum management are sufficient to address the resulting challenges and opportunities.
Broadband wireless services are poised to bring significant
benefits to all parts of the world. These new services will provide access to the internet and to IP-enabled services. Use of
the term Broadband Wireless Access or BWA really describes the
arrival of an era in which it is possible to enjoy the “internet
everywhere, all the time.” Because many BWA applications will
be ideal for consumers, this chapter will focus on frequencies
Spectrum regulators must also consider a growing number
of spectrum management “best practices” that have become
apparent over the last two or three decades. The best practices
explored in this chapter have fostered the widespread deployment of wireless services, including cellular mobile radio,
broadcast television, paging, and satellite services. Following
these practices has led to significant reductions in the cost of
C HAPTER 5
75
Figure 5.1: From Silos to Layers
The ultimate consumer broadband experience will be based on a multi- platform IP network with BWA as its core.
providing services and has created opportunities for entrepreneurs to develop innovative applications for consumers.
5.2 The Economics of Broadband Wireless Access
It is important for regulators to understand something
about the economics of wireless access, which are now – after
two decades of wireless mobility – becoming much clearer. The
principal economic drivers of wireless systems are the availability and cost of spectrum, the cost of the end-user device,
and the acquisition and maintenance of the subscriber (ranging from network management to billing and customer service operations). For a BWA system to be successful, it must be
competitive across each of these categories. This demonstrates
to operators and end users alike that it represents a viable value
in the broadband marketplace, particularly when compared
with wire-line or satellite alternatives that require significant
capital expenditures to increase capacity or coverage area.
5.2.1
Spectrum as An Input
Spectrum use is defined by four parameters, of which
only two – power and bandwidth – are usually set by regulators. These parameters determine the capacity and coverage
that a particular spectrum band can deliver to an operator for
the provision of services to end users. System capacity and coverage essentially determine the number, the size and the cost of
the transmitters (including the supporting backhaul network)
that an operator needs to deploy in order to deliver the desired
set of services. These factors, in turn, establish the financial
return threshold of the operator wishing to deploy a competitive BWA network.
Spectrum in the range below 3GHz has propagation
characteristics that enable wide coverage areas and can more
easily overcome interference from foliage, buildings and other
obstructions using non-line-of-sight technologies such as
beam forming. The ability to provide non-line-of-slight services reduces the number of base stations required to provide
coverage in these bands. But experience drawn from mobile
services indicates that an increasing customer density requires
76
more base stations to meet the growing demand for network
capacity.
Ultimately, any viable BWA service will need to enhance
both coverage and capacity, but initially it is more important
to have greater coverage. It is unlikely that new BWA services
will be in such immediate demand that networks will have to
be designed to maximize capacity over coverage. Over time,
as consumer adoption increases, BWA systems will begin to
emphasize throughput and capacity by reducing cell sizes and
increasing the number of base stations.
There are a number of spectrum bands operating between
0.4 GHz and 5.5 GHz that could help foster the growth
of BWA, although each band represents necessary tradeoffs
between capacity and coverage. In identifying these bands for
new uses such as BWA, the general trend across the globe has
been to reclaim the bands from incumbent licensees that have,
or will have, the ability to deploy more spectrally efficient
equipment that reduces their need for spectrum.
One category of reclaimed spectrum comes from government and military operations (1.5-2.4 GHz as well as 5.1-5.8
GHz). In the commercial context, satellite systems (2.0-2.3
GHz) and fixed microwave systems (3.1-3.7 GHz and 2.12.2 GHz) provide an avenue for reclaiming spectrum. The
increasing spectral efficiency of the latest technologies, plus the
widespread availability of substitute technologies and services
(for example, submarine cables and fiber-optic networks) has
reduced the demand for spectrum needed to deploy these services. Similarly, broadcast television bands – including the UHF
(400-700 MHz) band – and Multi-channel Multipoint Distribution Service (MMDS) bands (2.5 -2.7 GHz) have also been
the source for new BWA spectrum in various markets, because
newer, more spectrally efficient digital television broadcast
standards have been developed.
It is relatively easy to identify bands for new uses such as
BWA, especially in the context of global harmonization trends
that can provide clues to regulators about the most likely future
use of a spectrum band. The difficulty lies in transferring these
bands from an incumbent licensee to a new one. This paper
seeks to help address these transitional issues in a pragmatic
C HAPTER 5
Box 5.1: Defining Broadband Wireless Access
Internationally, the ITU considers broadband wireless access (BWA) to encompass mobile or fixed access technologies that
provide connections at speeds higher than the primary rate (e.g. 2 Mbit/s). This encompasses technologies within the IMT-2000
family, as well as newer technologies such as WiMax and WiBro. Nonetheless, it is likely that each country will continue to make
its own decision about the definition of BWA given its technological and economic development stage.
Singapore’s regulatory authority describes wireless broadband as “an access technology that offers high-speed data access
over the air. A wireless broadband network, typically operating at frequency bands less than 6 GHz, provides broadband speeds
ranging from 256 kbit/s to tens of Mbit/s. Each base station generally serves an area of up to several square kilometres. Wireless
broadband networks can deliver network connectivity to fixed locations using standards like IEEE 802.16d, and in the near future,
to mobile users using standards like IEEE 802.16e and IEEE 802.20.”
Kenya’s Communications Commission defines broadband fixed wireless access “as intentional radiators that use wideband
digital modulation techniques and provide a wide array of high data-rate fixed communications for individuals, businesses, and
institutions.”
Mauritius uses a three-part definition for BWA, in accordance with ITU-R Recommendations:
• Wireless Access systems are broadband radio systems that may be deployed either indoors or outdoors. These systems include:
– Fixed wireless access which may be defined as “Wireless access application[s] in which the location of the end-user
termination and the network access point to be connected to the end-user are fixed.”
– Mobile Wireless Access which may be defined as “Wireless access application[s] in which the location of the end-user
termination is mobile.”
– Nomadic Wireless Access which may be defined as “Wireless access application[s] in which the location of the end-user
termination may be in different places but it must be stationary while in use.”
Box 5.2: Spectrum Bands for BWA
450 – 500 MHz (Regions 2 and 3) and 600-1000 MHz (Region 1) – These could be used for broadband mobile access services,
given the significantly enhanced propagation characteristics of the band. These bands are under discussion as candidates to be
included as IMT-2000 bands during the 2007 ITU World Radiocommunication Conference (WRC).
1.5-2.5 GHz – Many of the bands in this range are currently identified as IMT-2000 bands. They are also under discussion,
across all regions, as part of the 2007 WRC process. They could be used for both fixed BWA and broadband mobile services sometimes referred to as “IMT-2000 and beyond” (that is, “3.5G” and “4G” technologies). These bands are also identified for hybrid
satellite and terrestrial wireless systems that permit seamless continental roaming for BWA operators.
3.4-3.7 GHz – Allocated across all regions for licensed BWA services.
5.1-5.7 GHz – These are allocated across all regions for unlicensed BWA uses.
fashion by creating marketplace incentives that encourage
licensees to transfer spectrum to its best and highest use, while
at the same time offering services that are competitive for the
particular dynamics and context of each country (whether rural
or urban, developed or developing markets).
5.2.2
End User Costs and Devices
End-user take-up of wireless services is largely a function
of the cost of the end-user device and of the applications that
work on that device. Common sense says that the lower the
costs of the handset, the more likely consumers are to adopt
the underlying service.
Over the last 20 years – but increasingly over the last five
years – the mobile marketplace has demonstrated that consumer adoption in developed markets has accelerated when
device costs have been below the USD 200 barrier. Although
C HAPTER 5
this threshold amount might not be practical for developing
markets, it is important to understand that mass adoption of
end-user devices in developed countries has the positive scale
effects that can easily translate to lower costs for end-user
devices distributed in developing economies. Additionally,
wireless equipment manufacturers have also learned from the
mobile market and are now offering end-user devices with
limited functionality to developing country markets in order to
further accelerate the manufacturing scale effect that reduces
the overall cost of making these devices.
5.2.3
Service Delivery and Management
Reducing the cost and complexity of applications, and the
associated service delivery mechanisms, that work on wireless devices has a significant effect on consumer adoption and
leads to accelerating the scale economies of manufacturing the
77
devices. In all geographic markets, the advent of flat-rated voice
services and the use of prepaid pricing plans have increased
the rate of adoption of wireless mobile devices and led to significant economies of scale. Additionally, the drive to add IP
services to wireless devices will increase the rate of consumer
adoption by making wireless access at the service layer indistinguishable from wired alternatives.
The impact of the IP on BWA service delivery can be best
understood by looking at the IEEE 802.16 WiMAX standards
development process currently under way. The key innovation of the 802.16 process has been to simplify the communications protocol stack, so that economies of scale can take
place in developing radios and the associated chipsets – the
most expensive elements – while providing greater freedom for
developing applications at the service layer.
5.3 The Technology Revolution
In addition to understanding the economics of spectrumbased services, regulators should also have a working understanding of the revolutionary changes in technology – many of
them driven by micro-electronics – that are permanently altering radio systems and spectrum engineering.
5.3.1
Applied Information Theory in Radio System
Design
Spectrum represents the temporal and spatial opportunities to transmit information using the electromagnetic
spectrum. The range of frequencies in the electromagnetic
spectrum is typically divided into eight bands, spanning from
3 Hertz to 300 GHz.
The characteristics of signal propagation depend on the
frequency band on which the signal is transmitted. These signals are typically transmitted by an antenna device that transmits energy in one or multiple directions. Shannon’s capacity
theorem, the fundamental theorem of radio communications
design, states that the rate of information transfer by a radio
is limited by the available bandwidth and the ratio of signal to
noise within the band. Within this physical limit, the spectrum
resources available to a radio are determined by four factors:
(1) specified bandwidth, (2) the allowable power or energy
emission within the band, (3) the bit error rate acceptable to
the end user and (4) the throughput desired by the consumer.
Most regulators have, up until now, only defined the
power limits and the bandwidth and left the other two factors to be determined by the marketplace. Traditional radio
systems were designed using analogue equipment and had
limited computational power available to them. The results
were radios that operated using very narrow throughput and
bit error rate (that is, quality-of-service demands) parameters
and that were highly sensitive to the operating environment.
Traditional radios were not flexible and could not be used to
take on new tasks or operate in new environments. This made
them unsuitable for broadband networks.
78
5.3.2
Advances in Microprocessor Technology
The advent of miniaturized and powerful computing resources available through digital signal processors (DSPs),
non-programmable hardware computing components and
field programmable gate arrays (FPGAs) has made it possible to
create radio systems that can dynamically change along all four
vectors (bandwidth, power, throughput, and bit error rate)
that define spectrum. By using powerful microprocessors to
dynamically change the four variables, radio system designers
can now create new spectrum capabilities where none had previously existed (see Figure 5.3). By incorporating digital and
microprocessor technologies into the design fabric of radio systems, engineers are now creating radios that can dynamically
operate outside the constraints of a particular intersection of
bandwidth and power limits normally set by regulators.
For example, a radio designed to optimally perform at
design PB (in the middle of the circle in Figure 5.3) can now
be redesigned, on the fly, to operate within any of the five
possible regions surrounding the optimal point. Each of the
five regions represents a trade-off between power, bandwidth,
throughput and bit error rate. The light areas (those below
the design line) represent increased performance in the form
of either throughput or bit error rate, in exchange for reduced
power and bandwidth. In contrast, the darker areas above the
design line provide for increased power and bandwidth utilization but poorer performance in terms of throughput and bit
error rate. Computing resources enable radios to make these
impromptu tradeoffs, increasing the flexibility of these systems to handle different types of market environments without
having to change radios.
The net result is that consumer radios are becoming
more flexible and highly adaptable. From a spectrum regulator’s perspective, the additional flexibility and adaptability of
the technologies means that new spectrum resources must be
accounted for and usage rights must be assigned.
5.4 Adapting Spectrum Regulatory Models for BWA
Keeping in mind the issues of spectrum economics and
technological advances, spectrum managers in the new broadband era face the challenge of achieving three separate but
interrelated goals, simultaneously:
(1) To provide the proper incentives for spectrum licensees,
both existing and new, to invest in broadband services;
(2) To expand consumer choices by enabling sustainable competition for similar services across multiple technological
platforms; and
(3) Implementing policies that discourage wasteful and anticompetitive behaviour resulting from uneconomic speculation and hoarding of spectrum.
5.4.1
Existing Regulatory Models
There are essentially three conventional methods of managing spectrum to attempt to achieve these goals. First, there
is the command and control model, in which strict operating
parameters and service rules define licensees’ spectrum rights.
C HAPTER 5
Figure 5.2: From VLF to EHF
The ultimate consumer broadband experience will be based on a multi- platform IP network with BWA as its core.
Frequency Range
Description
3 to 30 Khz
Very low frequency band (VLF)
30-300 KHz
Low frequency band (LF)
300 KHz-3 MHz
Medium frequency band (MF)
3 MHz-30 MHz
High frequency band (HF)
30 MHz-300 MHz
Very high frequency band (VHF)
300 MHz-3 GHz
Ultra high frequency band (UHF)
3 GHz-30 GHz
Super high frequency band (SHF)
> 30 GHz
Extremely high frequency band (EHF)
Figure 5.3: Changing the Technology Paradigm
New technologies permit new spectrum uses to be created by trading off power or bandwidth with throughput and bit error rate dynamically on the same radio.
An alternative licensing model is the exclusive rights model, in
which a licensee is given rights – which may be (within limits)
transferable and flexible – to use a specified spectrum band
within a defined geographic area and during a fixed period
of time. In the current understanding of the exclusive rights
model, spectrum use rules are primarily technical (as opposed
to service-based), because they are designed only to protect the
spectrum licensee and adjacent spectrum users from generating or receiving harmful interference – not to mold or develop
a certain service or market structure.
The third regulatory model is the commons model, or unlicensed model, which allows unlimited numbers of users to
share a block of frequencies without giving any one user or
group of users priority or individualized rights of use. Uses are
limited only by technical criteria that specify bandwidth and
C HAPTER 5
emitted power but provide no enforceable rights to protect
against interference. A well-known form of commons approach
has been the deployment of WLANs using Wi-Fi technology.
Having defined these models, it must be said that none
of them is adequate by themselves to address today’s rapidly evolving world of broadband spectrum. The command
and control approach grants spectrum rights on such narrow
grounds that they are often of limited utility for broadband
opportunities. Plus, the model requires constant government
intervention, with attendant delays and hassles, to change the
operating rules of the licences. The exclusive use model has
compelling arguments for providing market incentives to new
entrants. But it also creates perverse incentives for incumbent
licensees to engage in speculative or anti-competitive hoarding of spectrum, as a way to thwart real or perceived competi-
79
Box 5.3: Software Defined Radios, Adaptive Array Systems and Mesh Networks
The availability of computing resources for radio signal processing has enabled three exciting advances in radio technology:
software defined radios (SDRs), Adaptive Array Systems (AASs) and Mesh Networks.
SDRs are essentially radios that can be re-configured and adapted at the point of use and for different applications. This results
in a multi-band, multipurpose radio. In the ideal software radio scenario, the radio signal is directly converted to digital signals
at the antenna. All other radio functions are performed in the digital domain by software on the host platform, which might be a
flexible digital signal processing (DSP) chip, a computer or even a mobile telephone.
Unlike conventional antennas, where the energy is diffuse, AASs use computational algorithms to direct energy to parallel
and simultaneous channels within the same frequency bands. This exciting technology uses computing to combine an array of
antennas and radio frequency energy in order to detect and calculate radio signals on a highly refined basis. This allows the system
to suppress interfering signals and automatically track desired signals. The result is a significant increase in overall system capacity
by enabling greater re-use of the same radio frequencies.
Mesh networks were developed through military communications technology research. Also known as ad hoc or “infrastructure-free” networks, they are designed to maintain high quality-of-service in unstructured or harsh spectrum environments. Mesh
networks eliminate the need for a spoke and hub radio network and allow one subscriber to communicate with another without
requiring synchronization from the base station (See Box 3.5 in Chapter 3). Because mesh networks lack any common infrastructure, they represent new spectrum capabilities that are ungovernable from a central point, making them the latest challenge to
regulators’ ability to manage and allocate spectrum use rights.
tion. Finally, the commons approach – which today provides
low entry barriers and can result in significant scaling in terms
of consumer adoption – can become self-limiting through the
lack of enforcement mechanisms to manage and prevent overcrowding and overuse.
Rather than adopting any of these models in isolation,
regulators can adapt and combine them in ways that are appropriate to the circumstances. Above all, regulators can condition
the granting of new spectrum resources or rights upon demonstrations that they will be used to increase overall competition
for broadband services across all platforms. Moreover, spectrum rights or access can be granted to entities that prove they
are capable and willing to use them economically, because they
are willing to experience the real and quantifiable opportunity
costs of these spectrum rights. This pragmatic approach insures
that market forces are harnessed to achieve the goal of deploying BWA systems as rapidly and as efficiently as possible.
5.4.2
Re-Examining the Models
At this juncture, regulators may wish to review the existing models to see how they might be updated or adapted to
address the newer technological and market realities of the
BWA era. The following subsections consider each model in
turn.
5.4.2.1
The Command-and-Control Model
Some experts view the command-and-control model
as simply a means to exercise tight government control over
spectrum use. A more honest assessment, however, is that the
approach grew out of the realities of conventional radio design
during the last century. Essentially, those realities called for four
steps, all of them bureaucratic: (1) allocation, (2) enactment of
service rules, (3) assignment (licensing), and (4) enforcement
of the rules and license requirements.
80
The regulator’s task is never done under a command-andcontrol regime. Regulators must continually revisit and referee
the spectrum environment as new radios are introduced into
the marketplace. In order for a new radio system to enter the
market, the regulator must address everything from system
configuration, co-channel and adjacent channel effects, power
flux density, coding, out-of-band emissions, and innumerable
other technical criteria. Of course, the cost of the regulator’s
involvement is the time it takes to transfer spectrum rights to
their best possible uses. Given the rapid rate of technological
change, the time lags associated with government allocation
processes amount to a significant drag on the ability of licensees to rapidly deploy new BWA systems.
5.4.2.2
The Exclusive Use Model
In essence, this model gives licensees the right to use their
spectrum however they see fit, as long as they follow technical
rules that are designed to protect the licensee from causing or
receiving interference. In its purest form, the model confers
something like a fee-simple “ownership” of spectrum (with
technical covenants) for a set period of time. Licensees face few
or no restrictions on marketing their spectrum rights, including through secondary market trading of the usage rights.
Proponents of the exclusive licensing approach tout the
economic incentives it provides for licensees to seek the most
productive and profitable use of spectrum. Detractors, however,
fear that the model could lead to hoarding or under-investment
by incumbent providers that seek to raise costs for competitors
who need spectrum as an input to their own offerings. Without competition, there may be few incentives for incumbents
to develop new services or more spectrum-efficient systems.
Under the exclusive use model, incumbents left to their own
devices can simply buy up spectrum rights, with no guarantee they will use those resources to advance innovative and
C HAPTER 5
competitive wireless systems (including BWA networks) in the
marketplace.
5.4.2.3
The Spectrum Commons Model
The benefit of the commons model is that it removes the
input costs of obtaining spectrum from the economic equation
of market entry. Service providers can begin service at lower
costs, then scale up rapidly. End users face lower service prices
and can more easily afford handsets and terminal devices. Regulators wanting to jumpstart BWA deployment might easily
select the commons approach, especially in rural and underdeveloped areas where factors such as low population density and lack of backhaul infrastructure might otherwise limit
broadband service deployment and take-up.
Of course, the commons model has risks. Regulators
must consider the potential long-term effects of reducing entry
barriers and eliminating licensing. Commons models hold the
keys to their own undoing, because the very success in rapidly
deploying systems and proliferating users can lead to interference, crowding and an unstable spectrum environment. This
overcrowding phenomenon is more commonly known as the
“tragedy of the commons.” Regulators can mitigate the risk by
implementing power limits, modulation requirements, backoff schemes, and other measures. Of course, imposing new
technical rules by regulatory fiat might effectively convert the
commons approach into something more closely resembling
the classical command-and-control model, defeating the very
basis for establishing a commons model in the first place.
Some regulators are using a mix of licensed and unlicensed spectrum to address the need for low-cost broadband
services in rural areas. Ireland, for example, allows small operators to launch services in rural areas using unlicensed spectrum.
When the operators establish a successful business case, they
can migrate to licensed spectrum (See Box 5.4).
5.4.3
A Flexible Approach for New Times
In the broadband context, the fundamental future challenge for spectrum regulators is how to efficiently and effectively distribute new spectrum resources that technology is
making available. A pragmatic model, unencumbered by any
particular spectrum theology but able to draw from them all,
seems the optimal way to meet this complex challenge.
This pragmatic approach takes into account the fact that
technology is creating new spectrum capabilities and resources
faster than most regulators can redistribute spectrum rights
using traditional means. The modern spectrum regulator needs
a practical, outcome-oriented policy framework. The desired
result, in this case, is rapid deployment of broadband services, not a grand theory about the rights involved in spectrum
use. The problem with the traditional spectrum management
models is that they focus on defining usage rights of spectrum
licensees without articulating how to help achieve the underlying policy goals of the regulator.
Today’s broadband markets feature both high levels of
spectrum licence incumbency and low levels of inter-modal
competition. A pragmatic approach that rewards economic risktaking by spectrum holders will reduce the likelihood that they
C HAPTER 5
will “warehouse” spectrum simply for its perceived scarcity
value.
So what constitutes a pragmatic approach? Regulators can
start by offering to grant spectrum holders maximum flexibility for their spectrum rights on the condition they meet two
threshold obligations:
(1) They must demonstrate, even before gaining any new
spectrum rights, their commitment to increasing intermodal broadband competition; and
(2) They must agree to license conditions that positively
enforce the opportunity cost of their newly allocated spectrum rights.
5.4.3.1
Defining Flexible Spectrum Rights
In this pragmatic model, flexibility is defined as giving
licensees enough freedom to respond to market signals and
demand, within technical parameters designed to avoid harmful interference with adjacent licensees. As the universe of
wireless services expand, licensees must use these newfound
capabilities to provide new capabilities that will be demanded
from broadband networks. These capabilities include:
•
Providing either portable or mobile services to increase the
personalization of communication services and enhance
societal and individual productivity;
•
Achieving spectral efficiency and overall network efficiency to create economic returns on investments;
•
Reducing the cost of the customer equipment (handsets
and terminals) in order to encourage rapid and widespread
consumer acceptance; and
•
Enabling integration and convergence with other platforms to provide seamless connectivity.
With the advent of new technologies, licensees can make
tradeoffs between power, bandwidth, throughput and bit error
rate, in order to provide viable offerings in the broadband
marketplace. Furthermore, flexible spectrum rights will allow
licensees to apportion and share spectrum access with others,
facilitating the availability of BWA services and increasing competition.
5.4.3.2
Creating Competitive Broadband Markets
A practical approach to BWA spectrum licensing calls for
granting spectrum licensees not only technical flexibility to
create more spectrum capabilities and resources, but also operational autonomy to enter new lines of business whenever technology allows them to. Regulators should grant autonomy to
enter allied or new markets as long as doing so would enhance
competition and would enable the licensee to make the fullest,
most efficient economic use of the spectrum.
If the spectrum regulator simply grants technical and
operational flexibility, there is no guarantee that the licensee
will choose to deploy services that will add to competition. But
granting flexibility with a concomitant obligation to provide
competitive inter-modal broadband services would encourage
licensees to enter and aggressively compete in emerging markets, such as the BWA market.
81
Box 5.4: Eire’s Response to BWA
Ireland contributed several “principles” to the 2005 GSR Best Practice Guidelines on Spectrum Management to Promote Broadband Access.
Here are excerpts, discussed as “Principle Two” and “Principle Three.”
Principle Two: Balancing the Use of Licensed and Licence-Exempt Spectrum.
A number of local initiatives have taken effect to provide broadband access using licence-exempt spectrum. In Ireland, from
July 2002, wideband data transmission systems for the provision of fixed wireless access networks/metropolitan area networks
(FWA/MAN) have been permitted in the 5.8 GHz (5725 – 5875 MHz) band on a licence-exempt basis, provided that the maximum
radiated power does not exceed 2W eirp. This higher power level, over and above the current European harmonized standard, has
increased the coverage achievable and hence the utility of the 5.8 GHz band.
This initiative provided some impetus for small market players to enter the market at very low cost, to gain some experience
of broadband provision and to test-market demand for various broadband services. A number of successful operations using the
licence-exempt spectrum, having proved their business case, have now moved to licensed spectrum. [Irish regulator] ComReg has
committed itself to continue to identify appropriate spectrum allocations, both licensed and licence-exempt, for Wireless Access
Services that are supported by choice and availability of equipment.
Principle Three: Access to Cost Effective Backhaul Infrastructure.
Just as consumers in semi-rural or rural areas may not have access to ADSL, the providers of wireless broadband are hampered
by the lack of cost-effective backhaul infrastructure, e.g., fibre. The alternatives such as satellite or point-to-point wireless fixed
links are significantly more expensive compared to the costs of providing a wireless base station for broadband access.
In Ireland, consideration is being given to permitting the use of point-to-point links within the broadband access spectrum
to provide a cost effective backhaul operation. While this is difficult to accomplish from a spectrum management viewpoint, it is
seen as a viable alternative to the traditional and more expensive alternatives.
Ireland’s contribution to the 2005 GSR Best Practice Guidelines is available at http://www.itu.int/ITU-D/treg/Events/
Seminars/2005/GSR05/consultation.html.
5.4.3.3
In addition to promoting competition, flexibility should
be harnessed to enforce on the licensees the opportunity cost
of using spectrum. Generally, the goal here is to give licensees
price signals about the value of their spectrum holdings that
discourage them from engaging in uneconomic hoarding of
spectrum. Although regulators can combat hoarding by simply
recapturing spectrum, there are other, more productive mechanisms for enforcing the opportunity cost of spectrum.
Singapore’s distribution of spectrum for BWA services was
conducted in open and transparent fashion. IDA had earmarked
the 2.3 and 2.5 GHz bands for wireless broadband services in
February 2004. The following April, IDA launched a public
consultation proceeding on spectrum allocation and the licensing framework for wireless broadband services. IDA released
licensing details for broadband wireless services in February
2005, notifying interested parties that it would hold an auction
if demand exceeded the supply of available spectrum.
One of the more popular methods in the last decade has
been the use of auctions. Unfortunately, the effectiveness of
auctions has diminished as a result of the growth of the wireless industry and the availability of large amounts of capital.
This has desensitized the industry to price signals coming from
standard auctions. Furthermore, in some cases auctions have
been used to increase national revenues rather than as a tool for
enforcing market discipline. This has bred some unsustainable
auction results, creating uncertainty in the markets. Nonetheless, adjusting for these two factors, transparent auction processes are a viable method of enforcing the opportunity cost of
using spectrum.
Enabling secondary markets for trading spectrum rights
also has the effect of enforcing the opportunity cost of spectrum. Allowing the rapid transfer of spectrum rights between
private parties that value these rights differently creates price
signals that encourage licensees to use the spectrum to provide
competitive BWA services, because these uses will be the most
valued in the current marketplace. It is important to remember
that BWA services are the core component of a general set of
broadband services that consumers would utilize. Providing
the marketplace the flexibility to combine BWA services with
other platforms is more likely to increase consumer welfare
than by restricting such combinations.
In May 2005, for example, the Info-Communications
Development Authority (IDA) of Singapore successfully auctioned spectrum in the 2.3 GHz and 2.5 GHz frequency bands
for broadband wireless access services. The starting price for
each of the spectrum blocks put up for auction was SGD
1,000, and the highest closing price bid was SGD 550,000. IDA
decided to grant successful bidders a 10-year licence, in order
to provide investment certainty.
Additionally, with secondary markets for spectrum rights
gaining greater acceptance among licensees and regulators,
experts in the field are now considering how to combine auctions and secondary markets to create new mechanisms that
rapidly drive spectrum to its highest and best use. Two-sided
auctions, for example, let regulators and spectrum incumbents
combine their spectrum resources into a simultaneous auction
that transparently recalibrates both the geographic and tech-
82
Enforcing Opportunity Costs
C HAPTER 5
nical limits on spectrum rights. Two-sided auctions are being
developed as a way of smoothly restructuring bands to allow
new, innovative services like BWA.
• Adopting harmonized frequency plans defined by
ITU-R recommendation in order to facilitate the implementation of competition.
Another more traditional, but equally efficient, method of
enforcing opportunity cost is to impose build-out or construction obligations on licensees. Although these build-out obligations are effective in imposing costs, they tend to be blunt
regulatory instruments because they are conditioned on prior
assumptions about marketplace conditions. But combined
with secondary spectrum markets, these obligations can serve a
valuable role in dissuading licensees from hoarding spectrum.
• Embracing the principle of minimum necessary regulation, where possible, to reduce or eliminate regulatory
barriers to spectrum access, including simplified licence
and authorization procedures for the use of spectrum
resources.
In shared license bands, the sharing rules developed by
regulators are the best method of enforcing the opportunity
cost of using the spectrum. These rules generally determine
the level of barriers to entering the shared bands, the amount
of interference permitted between and among users of the
band, and the power levels permitted (and therefore the coverage range). These key parameters all help to define the opportunity cost of using the spectrum. In the case of unlicensed
bands, low power limits combined with the lack of interference
management circumscribe the use of the bands and encourage
a high degree of efficiency.
5.5 Defining Best Practices
Spectrum regulators need also to look at a number of
spectrum management best practices developed over the last
two or three decades as tools that can be used to encourage
BWA deployment. This section explores several versions of
best practices, beginning with those specifically endorsed by
the world’s regulators, as members of the Global Symposium
for Regulators (GSR).
5.5.1
The GSR Guidelines
Recognizing that spectrum is a scarce resource that needs
to be managed effectively and efficiently, the delegates to the
2005 GSR, held in Yasmine Hammamet, Tunisia, drafted a
set of best practice guidelines for spectrum management to
promote broadband access. The 2005 guidelines continue the
tradition of best practices agreed to at the GSR conferences
in 2003 and 2004 on promotion of universal access, and lowcost broadband services, respectively.2 The 2005 guidelines are
reprinted here in full:
“We, the regulators participating in the 2005 Global Symposium for Regulators, have identified the following:
1.
Facilitate deployment of innovative broadband
technologies: Regulators are encouraged to adopt policies to promote innovative services and technologies. Such
polices may include:
• Managing spectrum in the public interest.
• Promoting innovation and the introduction of new
radio applications and technologies.
• Reducing or removing unnecessary restrictions on
spectrum use.
C HAPTER 5
• Allocating frequencies in a manner to facilitate entry
into the market of new competitors.
• Ensuring that broadband wireless operators have as
wide a choice as possible of the spectrum they may
access, and releasing spectrum to the market as soon as
possible.
2.
Promote transparency: Regulators are encouraged to
adopt transparent and non-discriminatory spectrum management policies to ensure adequate availability of spectrum, provide regulatory certainty and to promote investment. These policies may include:
• Carrying out public consultations on spectrum management policies and procedures to allow interested
parties to participate in the decision-making process,
such as:
public consultations before changing national frequency allocation plans; and
public consultations on spectrum management decisions likely to affect service providers.
• Implementing a stable decision-making process that
provides certainty that the grant of radio spectrum is
done in accordance with principles of openness, transparency, objectivity – based on a clear and publicly
available set of criterion which is published on the
regulator’s website –and non-discrimination and that
such grants will not be changed by the regulator without good cause.
• Publication of forecasts of spectrum usage and allocation needs, in particular on the regulator’s website.
• Publication of frequency allocation plans, including
frequencies available for wireless broadband access, in
particular on the regulator’s website.
• Publication of a web-based register that gives an overview of assigned spectrum rights, vacant spectrum, and
licence-free spectrum, balancing any concerns for confidential business information or public security.
• Clearly defining and publishing radio frequency spectrum users’ rights and obligations, including on the
regulator’s website.
• Clearly defining and publishing licensing and authorization rules and procedures, including on the regulator’s website.
• Publication of legal requirements for imported equipment and foreign investment, in particular on the relevant government agency website.
83
3.
4.
84
Embrace technology neutrality. To maximize innovation, create conditions for the development of broadband
services, reduce investment risks and stimulate competition among different technologies, regulators can give
industry the freedom and flexibility to deploy their choice
of technologies and decide on the most appropriate technology in their commercial interest rather than regulators specifying the types of technologies to be deployed,
or making spectrum available for a preferred broadband
application, taking into consideration the need for and
cost of interoperable platforms.
• Regulators can take into consideration technological
convergence, facilitating spectrum use for both fixed
and mobile services, ensuring that similar services are
not subject to disparate regulatory treatment.
• Regulators can provide technical guidelines on ways to
mitigate inter-operator interference.
• Regulators can ensure that bands are not allocated for
the exclusive use of particular services and that spectrum allocations are free of technology and service constraints as far as possible.
Adopt flexible use measures: Regulators are encouraged to adopt flexible measures for the use of spectrum for
wireless broadband services. Such measures may include:
• Minimizing barriers to entry and providing incentives
for small market players by allowing broadband suppliers to begin operations on a small scale at very low cost,
without imposing onerous rollout and coverage conditions, to enable small market players to gain experience
in broadband provision and to test market demand for
various broadband services.
• Recognizing that wireless broadband services may be
used for both commercial and non-commercial uses
(e.g., for community initiatives or public and social
purposes) and that broadband wireless spectrum can be
allocated for non-commercial uses with lower regulatory burdens, such as reduced, minimal or no spectrum
fees; regulators can also allocate and assign spectrum
for community or non-commercial use of broadband
wireless services.
• Recognizing through flexible licensing mechanisms
that wireless broadband technologies can provide a full
range of converged services.
• Adopting lighter regulatory approaches in rural and less
congested areas, such as flexible regulation of power
levels, the use of specialized antennas, the use of simple
authorizations, the use of geographic licensing areas,
lower spectrum fees and secondary markets in rural
areas.
• Recognizing that in markets where spectrum scarcity is
an issue, the introduction of mechanisms such as secondary markets can in some cases foster innovation and
free up spectrum for broadband use.
• Recognizing the role that both non-licensed (or licenceexempt) and licensed spectrum can play in the promotion of broadband services, balancing the desire to
5.
6.
7.
8.
9.
foster innovation with the need to control congestion
and interference. One measure that could be envisaged
is, for example, to allow small operators to start operations using licence-exempt spectrum, and then move
to licensed spectrum when the business case is proved.
• The promotion of shared-use bands, as long as interference is controlled. Spectrum sharing can be implemented on the basis of geography, time or frequency
separation.
• Developing strategies and implement mechanisms for
clearing bands for new services as appropriate.
• Recognizing the need for cost-effective backhaul infrastructure from rural and semi-rural areas, regulators
can consider the use of point-to-point links within
other bands, in line with national frequency plans,
including any bands for broadband wireless access.
Ensure affordability. Regulators can apply reasonable
spectrum fees for wireless broadband technologies to
foster the provision of innovative broadband services at
affordable prices, and minimize unreasonable costs that
are barriers to entry. Higher costs of access to spectrum
further reduce the economic viability in rural and underserved areas. Auctions and tender processes can also be
managed to meet these goals.
Optimize spectrum availability on a timely basis.
Regulators are encouraged to provide effective and timely
spectrum use and equipment authorizations to facilitate
the deployment and interoperability of infrastructure for
wireless broadband networks. Regulators are also encouraged to make all available spectrum bands for offer, subject to overall national ICT master-plans, in order that
prices are not pushed up due to restrictive supply and
limited amount of spectrum made available and so that
opportunities to use new and emerging technologies can
be accommodated in a timely manner. In addition, special
research or test authorizations could be issued to promote
the development of innovative wireless technologies.
Manage spectrum efficiently. Spectrum planning is
necessary to achieve efficient and effective spectrum
management on both a short-term and long-term basis.
Spectrum can be allocated in an economic and efficient
manner, and by relying on market forces, economic incentives and technical innovations. Regulators can promote
advanced, spectrum-efficient technologies that allow coexistence with other radio communications services, using
interference mitigation techniques (for example, dynamic
frequency selection). Regulators can provide swift and
effective enforcement of spectrum management policies
and regulations.
Ensure a level playing field. To prevent spectrum
hoarding, especially by incumbents, regulators can set
a limit on the maximum amount of spectrum that each
operator can obtain.
Harmonize international and regional practices and
standards. Regulators can, as far as practicable, harmonize
effective domestic and international spectrum practices
and utilize regional and international standards whenever
C HAPTER 5
Figure 5.4: Globally Harmonized Spectrum: IMT-2000
possible, and where appropriate, reflect them in national
standards, balancing harmonization goals with flexibility
measures. This could include harmonization of spectrum
for broadband wireless access that could generate economies of scale in the production and manufacture of equipment and network infrastructure. Likewise, global harmonization of standards to ensure interoperability between
different vendor’s user terminals and network equipment
can be promoted. The use of open, interoperable, nondiscriminatory and demand-driven standards meets the
needs of users and consumers. Coordination agreements
with neighbours, both on a bilateral or multilateral basis,
can hasten licensing and facilitate network planning.
10. Adopt a broad approach to promote broadband
access. Spectrum management alone is inadequate to
promote wireless broadband access. A broad approach,
including other regulatory instruments; such as effective
competitive safeguards, open access to infrastructure, universal access/service measures, the promotion of supply
and demand, licensing, roll-out and market entry measures; the introduction of data security and users’ rights,
where appropriate; encouraging the lowering or removal
of import duties on wireless broadband equipment; as
well as development of backbone and distribution networks is necessary.”
5.5.2 Additional Best Practices
In addition to the GSR guidelines, it is useful to more
fully explore several best practices, to identify their benefits,
as well as their potential limitations, in advancing the cause of
consumer broadband services. This subsection will analyse the
following best practices:
• Harmonizing spectrum allocations on a global basis to
increase economies of scale at product and the service
layers and to reduce end-user costs.
• Fostering the use of standards-based technologies to
increase economies of scale.
• Allocating spectrum and developing technical rules that
encourage adjacent spectrum users to have compatible technical characteristics, as a way of limiting interference and
maximizing use of spectrum (“good neighbour” policies).
• In shared spectrum bands, encouraging or mandating technical standards that foster cooperative systems designed to
reduce harmful interference.
C HAPTER 5
•
Develop efficient and transparent licensing rules and processes that allow for restructuring of incumbent spectrum
bands in order to implement harmonization goals.
5.5.2.1
Global Harmonization of Allocations
Broadband spectrum regulators can significantly drive
down the costs of broadband wireless services and boost subscriber numbers by harmonizing spectrum allocations with
global practices. Harmonization allows equipment manufacturers to benefit from economies of scale by manufacturing
equipment for large pools of customers in multiple countries.
Of course, the ideal of global harmonization, as depicted
in Figure 5.4 (IMT-2000 bands) can lead to a significant
amount of incumbent dislocation – with the regulator feeling
the attendant political backlash. This is especially true in developed countries with a lot of entrenched incumbents, most of
which feel they have rights under the exclusive spectrum use
model. Harmonization can usually be achieved only through
transparent licensing systems and significant political willpower.
Needless to say, participation in the ITU and regional
organizations can help provide the economic and political
support to require relocation of incumbents from newly harmonized bands. For example, ITU-R Study Group 8 and its
Working Party 8F are currently exploring potential allocations
for BWA technologies, including WiMAX. Governments need
to participate in these ITU groups, as well as in the industryled groups that pioneer standards. This will give spectrum
managers advanced warning about the global direction of these
technological developments, so they can establish appropriate spectrum regulations before the standards become widely
deployed.
5.5.2.2
Fostering Standards-Based Technologies
In order to maximize their spectrum allocation decisions,
spectrum regulators must closely follow and support the norms
and recommendations of multiple standards-setting organizations, including: (a) the Institute of Electrical and Electronics Engineers (IEEE); (b) the European Telecommunications
Standards Institute (ETSI); (c) the Wi-Fi Alliance (for 802.11
products); and (d) the WiMAX Forum (for 802.16 products).
Similar to global harmonization, the effect of embracing
standards-based technology development is to significantly
reduce the cost of the devices by reducing the number of pro-
85
prietary components. This leads to faster adoption of the equipment and the associated services. There are many benefits to a
standards-based approach, but it is also important to remember
that it is a consensus-driven process. If not managed carefully,
the standards process can essentially commoditize innovation
and could even become a de facto form of economic regulation,
especially if regulators embed the standards into their regulatory systems.
Broadband regulators also must be mindful that a standards-based approach, with its natural tendency towards compromise, could potentially lead to suboptimal results. These
unintended “costs” must be part of a broader policy tradeoff in
embracing any particular standard. Generally, regulators should
avoid embedding standards in their policies and rules without
an extremely compelling policy rationale.
5.5.2.3
‘Good Neighbour’ Allocations
This best practice generally involves grouping spectrum
allocations based on interference and other technical compatibility characteristics. This good neighbour “zoning” practice
can enhance the compatibility of spectrum uses, based on
power or bandwidth characteristics, maximizing overall capacity and reducing transaction costs. In a fashion similar to the
global harmonization of spectrum allocations, this practice
also raises the potential for significant relocation costs, which
regulators have to consider in making any decision to realign
spectrum allocations.
5.5.2.4
Voluntary Sharing Guidelines for Unlicensed Bands
For users of shared spectrum – such as those in unlicensed
bands – regulators can encourage voluntary coordination to
better manage any interference or capacity issues. Users of
these bands should understand that the alternative to self-regulation might be reverting to a command-and-control model
that would deny them service and technology flexibility.
5.5.2.5
Infrastructure Sharing
Regulators can also enhance wireless broadband adoption
by inducing licensees to share infrastructure such as towers
and backhaul facilities. Infrastructure sharing between wireless systems promotes efficiency, reduces deployment costs and
reduces environmental impacts by avoiding the construction of
duplicative facilities.
5.5.2.6
Setting Different Power Limits for Rural Areas
Regulators should adopt different rules for spectrum used
in different geographic markets. In lower-density environments, such as rural areas or under-served communities, there
is less opportunity for interference. So it may make sense to
allow transmitters to operate at higher power levels. Similarly,
regulators can increase the size of the bandwidth allocated to
broadband services in these areas in order to increase capacity.
With limited competition for spectrum in these areas, granting flexibility might provide better economic incentives for the
licensee to deploy a BWA network. Ireland’s efforts to grant
rural BWA operators greater flexibility with regard to spectral power limits has enabled the country to meet the broad-
86
band demands of its rural population. In addition, Ireland has
endeavoured to keep licensing obligations as low as possible to
reduce barriers to entry (See Box 5.5).
5.5.2.7
Transparent Licensing Systems and Processes
Another best practice is for regulators to establish transparent and automated licensing procedures and records in order
to reduce transaction costs associated with facilitating the highest and best use of spectrum rights. As part of their licensing
efforts, regulators should periodically test, analyse, and audit
spectrum resources to measure how efficiently incumbents are
using it and whether there are interference issues. Regulators
can use this information to form improved sharing and interference rules and regulations, as well as to expand spectrum
capabilities and permit new forms of uses within the band for
either the incumbents or new licensees.
5.5.3 Technology Neutrality
Technology neutrality is usually understood to be when
regulators apply rules and regulations in a way that does not
favour one type of technology over another. Clearly, this concept could be at odds with some of the other best practices,
such as global harmonization, adopting standards-based technologies and applying good-neighbour allocation rules, which
by their very nature tend to favour particular technological
choices.
From the perspective of the regulator as a resource manager, full technological neutrality is an impossible goal, because
the desire to achieve efficiency and rapid utilization of the
spectrum ultimately requires decisions that point to particular
technology paths. Promoting a standard or a particular spectrum band or its configuration, directly or indirectly, obliterates
the notion of neutrality. Similarly, creating a harmonized spectrum band and associated service rules will also tend to favour
particular technologies. So there is an inherent contradiction
between the goal of technological neutrality and the function
of the spectrum regulator as a resource manager.
Fortunately, pragmatism once again comes to the rescue.
A spectrum regulator has to play different roles that lead to different conclusions about technological neutrality. On the one
hand, the spectrum regulator is a resource manager concerned
with optimizing the efficient use of a scarce resource. On the
other hand, the spectrum regulator is a policy advocate trying
to achieve social policy goals such as universal access/service
and reducing the Digital Divide.
A practical solution to the conundrum would apply technological neutrality only to the means used to achieve macro
policy goals, rather than the means for managing the spectrum
resource. Technological neutrality is paramount only for the
means applied to achieve broad social policies such as attaining
universal broadband access. To achieve that goal, any combination of available technologies and resources can be enlisted.
One example of effective balancing between the goal of technology neutrality and pragmatism can be found in the public
consultation process of the Office of the Telecommunications
Authority (OFTA) in Hong Kong, China (See Box 5.6).
C HAPTER 5
Box 5.5: Eire’s Response to BWA, Part Two
Ireland’s response in the process of compiling the 2005 GSR Best Practice Guidelines on Spectrum Management to Promote Broadband Access
also addressed the issue of reducing licensing obligations as well to lower market-entry barriers.
Principle One:
Barriers to entry should be as low as possible
Our [Ireland’s] experience indicates that regulators should minimize barriers to entry in this area by allowing broadband
suppliers to begin operations on a small scale, and not imposing onerous rollout and coverage conditions. Ireland has awarded
national licences in the past for broadband wireless access that incorporated rollout and coverage obligations. But none of the
licensees were able to make a viable business case and, consequently, rollout of services was less than satisfactory.
In 2004 ComReg announced a new scheme for the licensing of broadband fixed wireless access services in local areas. Each
local service area was defined by a 15 km radius circle from a base station, with an interference zone extending to a 30 km radius,
at the perimeter of which a certain field strength should not be exceeded in order to limit interference into adjacent areas. Since
its inception, 110 licences have been granted on a first-come, first-served basis. The success of this approach, compared with the
earlier attempt at national licences, is reflected in an increase of 43 per cent of customers in the last six months.
One of the key reasons for the success is that operators only take out licences for areas in which they are able to develop a viable business case and, as there is no national network rollout obligation, all attention is focused on the local area. Initial concerns
that rollout would only occur in urban areas (due to high population) have proved to be unfounded as small entrepreneurs and
local community groups have taken up the challenge to supply broadband access to many rural areas where ADSL is not available.
Current rollout is shown in Figure 5.5.
Figure 5.5: BWA Coverage Areas (Circles) in Ireland
5.6 Case Study: BWA Spectrum Allocation in Mauritius
In order to gather all of the threads together in a real-world
example, this chapter concludes with a comprehensive review
of a major BWA allocation decision by the Mauritius Information and Communication Technologies Authority (ICTA).
During the first half of 2005, ICTA made a series of decisions to establish the future course of BWA services in this
fast growing island economy. Although it is too early to know
the results of ICTA’s allocation decisions, a brief review of the
Authority’s processes and methodology reveals a regulator that
has adopted key best practices and made pragmatic tradeoffs
that should enable BWA to take root rapidly.
C HAPTER 5
Unlike many regulatory bodies, ICTA has a broadly
defined mandate that, pursuant to its 2001 charter, calls for
it not just to “manage” spectrum but to increase the reach
of information and communication services throughout the
country. Turning its attention to BWA offerings, ICTA undertook a transparent public consultation process and arrived at its
final decision within a commendable 180 days after its initial
report.
In starting the process, ICTA first identified the demand
for BWA offerings, while also recognizing the need to harmonize its allocation decisions with global allocation trends, in
order to take advantage of scale economies. ICTA also iden-
87
Box 5.6: OFTA’s Consultation on Broadband Wireless Access Licensing
In August 2005, the Office of the Telecommunications Authority (OFTA) in Hong Kong, China, issued its “analysis of comments received, preliminary conclusions and further consultation on a licensing framework for deployment of broadband wireless
access.” Comments on the further consultation were invited through 31 October 2005. The August consultation followed an initial
BWA consultation launched in December 2004 on whether BWA should be licensed in Hong Kong, and if so, when.
In the August consultation, OFTA expressed its view that BWA spectrum should be assigned in 2006, on a technology-neutral basis. Consistent with the technology neutrality principle – and having considered the respondents’ views – OFTA said it
was prepared to allow the deployment of any technology that conformed to recognized open standards for the delivery of BWA
services. Because BWA devices and equipment will be supplied competitively, OFTA considered it unlikely that end users would
have insufficient choices in the selection of BWA devices.
OFTA also expressed the view that, although BWA is currently being deployed as a fixed service, it should also allow mobile
services, once the technology is developed and cost-effective. OFTA therefore proposed that the scope of permitted services of
the BWA licences be restricted to fixed telecommunication services initially, but expanded to include full mobility services after 1
January 2008. Fixed telecommunication service will include the conventional wireless local loop services, plus “limited mobility”
offerings, which will not be able to hand off calls between cell sites until after 1 January 2008.
OFTA also said it would issue unified carrier licences in order to accommodate the trend toward fixed-mobile convergence,
since BWA can offer both. The new Unified Carrier Licence would be valid for 15 years – the same duration as for existing fixed/
mobile carrier licences. OFTA has made clear that BWA licensees will be expected to invest in, and roll out, infrastructure to provide public services. They will not be allowed to enter the market solely as services-based operators. The consultation document
may be accessed at http://www.ofta.gov.hk/en/report-paper-guide/paper/consultation/20050831.pdf.
tified key factors that have negatively affected past attempts
in Mauritius to develop BWA services, such as congestion in
the unlicensed 2.4 GHz band. ICTA found that operators in
that band were exceeding designated power limits, using the
licence-exempt systems for long-range transmissions – contrary to their design and purpose. It also noted the continuing and significant demand from ISPs for wireless spectrum to
deploy their services.
Additionally, ICTA took note of other countries’ decisions
to define certain bands for BWA uses, including 2.5-2.7 GHz,
3.3-3.5 GHz, 5.1-5.3 GHz and higher-powered unlicensed use
of 5.4-5.8 GHz. ICTA took into consideration the potential of
new standards such as WiMAX for deploying BWA services
integrated with computers and other ICT devices. The ICTA’s
ultimate decisions demonstrated a practical approach to resolving the various tradeoffs needed to deploy BWA networks
in the face of competing demands from operators and other
incumbent spectrum users. The following subsections describe
how ICTA resolved issues in each band.
5.6.1 The 5.4-5.8 GHz Band
ICTA determined that the presence of radar incumbents
in the 5.4-5.8 GHz band required postponing any BWA allocation decision in the band, even though this band had been
globally harmonized for higher-powered unlicensed operations through the WRC 2003 negotiations. The complexity
of the radar operations, coupled with their national defence
purpose, will require a more deliberate transition for this band.
This issue is similar to the difficulty that the United States has
experienced in implementing the necessary dynamic frequency
selection (DFS) systems needed to protect sensitive military
operations in that band in the United States.
88
5.6.2 The 2.4 GHz BWA License-Exempt Band
In reviewing the status of the 2.4 GHz unlicensed band,
ICTA came to the conclusion that the previous “commons”
model had led to overuse and overcrowding. It found that existing operators in the band tended to exceed the power limits,
which were set at 23 dam EIRP, in order to extend their ranges
and overpower interference. ICTA decided to set a distance
limitation for this band to discourage operators from using it
for longer distance applications than the WLAN networking it
was originally intended for. In its decision, ICTA took some
pragmatic steps to improve the functionality and longevity of
the band:
• Mandated use of the bands for applications not to exceed
500 meters;
• Mandated limiting the emitted power to 20 dam while
giving some transitory leeway for incumbent operators to
stay at the 23 dam until 2010; and
• Required new systems to register with ICTA, allowing it
to track the level of usage in the band and identify potential interference problems.
These provisions may well extend the useful life licenceexempt operations in the 2.4 GHz band. Moreover, the problems encountered in Mauritius underline the pitfalls associated
with the commons approach.
5.6.3 The 2.5-2.7 GHz BWA Band
This band previously had been allocated for MMDS.
ICTA’s decision harmonized the band with the IMT-2000
allocation, enabling three distinct types of BWA systems to
eventually operate in the band. Implicit in the decision was the
determination that the existing MMDS use was not as relevant
or promising as the potential BWA systems that could use the
C HAPTER 5
Figure 5.6: The 2.5-2.7 GHz BWA Allocation in Mauritius
band. ICTA did not, however, make an explicit determination
about the value of MMDS relative to BWA. Rather, ICTA took
its cue from the public consultation process, in which only
one respondent advocated the continued use of the band for
MMDS.
Other aspects of the ICTA’s decision about this band
were significant. First, it determined that only licensed operators could provide BWA services in this band. Second, it made
sure to create a channelization plan for the band that would
group similar systems together, instead of using the interleaving employed with broadcast bands. As a result, both TDD and
FDD wideband systems could be deployed to deliver BWA.
The channel sizes were also changed to 5 MHz from 8 MHz
segments, which are better suited for wideband systems such
as those being developed by the IEEE WiMAX groups.
ICTA also allocated 40 MHz (20 MHz for uplink operations and 20 for downlinks) for hybrid satellite and terrestrial
services. Finally, ICTA created a transition period for incumbent operators, making the band available for BWA services
only in January 2010. ICTA left the door open for voluntary
relocation, but with language suggesting that the timeline for
transition might be accelerated based on “market conditions.”
5.6.4 The 3.4-3.6 GHz BWA Band
The ICTA reallocated this band from primarily a fixed satellite service band to a band in which those operations will be
co-primary with terrestrial BWA operations. Fixed-link services were favoured in this band in order to provide protection
for incumbent VSAT operators. Despite this limitation, BWA
C HAPTER 5
advocates such as wireless internet service providers (WISPs)
were able to obtain the benefit of higher-powered use, since
ICTA allowed for 15 W EIRP systems to operate in the band. In
assigning channels in these bands, ICTA also decided it would
give priority to public operators (although not conclusively);
that licenses were required and that the permitted point-topoint and point-to-multipoint links must be registered. Finally,
ICTA allowed aggregation of multiple 25 kilohertz channels,
while requiring a minimum of 100 MHz separation to avoid
interference of duplex operations in the band.
5.6.5 The 5.150-5.350 GHz BWA Band
Consistent with WRC 2003 agreements, ICTA opened up
this band for mobile licence-exempt use of equipment consistent with IEEE 802.11 (Wi-Fi) standards. Given the challenges
presented by incumbent radar operations in this band, ICTA
determined that Wi-Fi in this band would be limited to indoor
use only. Additionally, in order to prevent overcrowding and
potential interference to incumbent operations, ICTA required
the equipment in this band to use dynamic frequency selection
(DFS). This is an automated mechanism that detects the presence of signals from other systems, notably radar systems, and
avoids co-channel operation. Equipment must also have transmit power control (TPC), a mechanism that regulates a device’s
transmit power in response to an input signal or a condition.
These capabilities must be certified as part of the equipment
registration and approval process that ICTA plans to establish
for this band.
89
The limitations imposed on the use of this band reflect
the tradeoffs ICTA believed were necessary to advance BWA
while accommodating the sensitive operations already existing
in the band. Both DFC and TPC systems are at early stages in
their development and have not yet fully demonstrated their
ability to protect incumbent operations. Nonetheless, they are
clearly part of the technological advances enabling the rapid
deployment of BWA services.
5.7 Conclusion
The ability of broadband wireless access networks to
improve our lives ultimately relies on the amount of spectrum
regulators make available for BWA. For the first time in the
relatively short history of spectrum management, however,
it appears that advances in technology – independent of the
actions of regulators – can increase spectrum capabilities and
resources. New technologies allow users to do more with the
same amount of spectrum and enable new uses for spectrum
not previously possible. As these advances become more widely
adopted and new spectrum resources become available, regulators must consider whether traditional approaches to spectrum
management are sufficient to address the resulting challenges
and opportunities.
A sensible approach for BWA spectrum licensing calls for
granting spectrum licensees unlimited technical flexibility (so
long as they avoid harmful interference to adjacent licensees)
to create more spectrum capabilities and resources. Licensees
1
2
should have enough operational autonomy to enter new lines
of business. The pragmatic regulator grants these additional
rights to licensees as long as they meet two absolute preconditions important to the development of communications markets: to increase competition for broadband communication
services and to experience the opportunity cost of using their
spectrum assignments.
While contemplating the appropriate regulatory model for
the evolving state of spectrum technology, spectrum regulators must keep in mind key best practice concepts that have
developed around spectrum management over the last three
decades. These best practices fostered the widespread adoption
and deployment of an earlier generation of wireless services.
The same best practices will help deploy BWA networks today,
increasing the welfare of consumers in the modern Information Society.
In reviewing recent comprehensive BWA allocations in
Mauritius, one can trace the outlines of the practical tradeoffs
that are necessary to achieve the broader policy goals of the
regulators. The lesson to be drawn from this type of complex
allocation decision is not whether any particular outcome is
better than any other. The lesson comes from better understanding the need to balance the demands of allocating new
spectrum rights and resources for new technologies, while
simultaneously advancing broader social and policy goals such
as benefiting consumers and boosting economic productivity.
This balancing act, after all, is the task entrusted to the modern
spectrum regulator.
Opportunity cost, as defined by WiKipedia, is a term used in economics, to mean the cost of something in terms of an opportunity foregone (and the benefits that
could be received from that opportunity), or the most valuable foregone alternative. http://en.wikipedia.org/wiki/Main_Page
See http://www.itu.int/ITU-D/treg/Events/Seminars/2003/GSR/WSIS-Statement.html and http://www.itu.int/ITU-D/treg/Events/Seminars/2004/GSR04/consultation.html
90
C HAPTER 5
6 VOIP AND REGULATION
Authors: Tracy Cohen, Independent Communications Authority of South Africa (ICASA); Olli Matila,
Finnish Communications Regulatory Authority (FICORA) and Russell Southwood, Southwood Consultants
Voice over Internet Protocol (VoIP) service is often viewed
as a “disruptive technology,” meaning that it has the potential
to drastically alter the status quo in the global telecommunication industry.1 In fact, all of the current market indications
show that IP networks and VoIP services will replace traditional PSTN networks and services. ITU expects that by 2008,
at least 50 per cent of international minutes will be carried
on IP networks, and many carriers will have all-IP networks.
Recent trends are certainly headed in this direction. For example, in the United States, residential VoIP subscriber numbers
rose from 150,000 at the end of 2003 to more than 2 million in
March 2005. U.S. subscribership is expected to exceed 4.1 million by the end of 2006, generating over USD 1 billion in gross
revenues for the year.2
ITU believes that much of the mobile traffic in the world
will become IP-based, as well, and that the introduction of
mobile VoIP will influence the shape of the mobile business
globally. Wireless-enabled VoIP offers the potential for cheaper
voice calling. Users of 3G networks can already use mobile
phones to make VoIP calls at cheaper data rates. The growth
of mobile VoIP will particularly affect the high-priced international roaming business.
Today, however, VoIP services have been implemented
unevenly around the world. Some countries have legalized
and allowed multiple providers, while others have completely
blocked the provision of VoIP services. Some countries have
only “grey market” VoIP providers. These varying approaches
reflect very different perceptions of VoIP in various parts of the
world. In some countries, VoIP is seen as an exciting technological development that offers cheaper calling for consumers.
But in others, it is seen as an unauthorized threat to the existing
order. At the international level, VoIP traffic is often described
as “by-pass” or “lost” traffic, but it is driving the development
of new service providers in both the developed and developing
worlds.
The advent of VoIP has brought new challenges for regulators. In developing countries, where the entrenched rights
of fixed-line operators are most protected, the main question
has been whether to legalize the introduction of VoIP. In more
mature and competitive markets, meanwhile, VoIP has raised
questions about what aspects of it should be regulated. The
countries that have legalized VoIP as part of a broader liber-
C HAPTER 6
alization of their markets have started to accumulate thinking,
experience and precedent in this area. But even in these countries, VoIP is a relatively recent development, and there is often
little consensus about how to regulate it.
This chapter examines how VoIP services will affect
future regulation. Because of the starkly contrasting global
perceptions of VoIP, however, it is difficult to present a unified approach to regulating VoIP. Instead, this chapter aims to
reflect regulatory experiences from a wide range of countries
that are grappling with the transition to VoIP. The sections of
this chapter are structured to answer both the broad and specific questions raised by VoIP services, including:
• The overall approach to regulating VoIP as a mainstream
service;
• How VoIP has changed telephony business models and
the various ways of classifying the services it has created;
and
• Other related issues frequently raised in connection with
VoIP, such as quality of service; network integrity; communication security and lawful interception.
6.1 VoIP: Regulatory Evolution or Revolution?
6.1.1 How VoIP is Changing Voice Business
Models
In liberalized telecommunication environments, it is
important to create regulatory frameworks that allow the
market to produce sustainable business models. VoIP is a powerful service innovation that has the potential to change how
existing voice markets operate. Since VoIP service is largely
enabled by the existence of IP networks, there is inevitably an
overlap throughout this chapter between these two key concepts, which remain inextricably linked. It is not yet clear how
IP networks will be implemented – or at what speed – but it
is important to try and identify key elements of the changing
business model in order to understand the policy and regulatory dilemmas VoIP raises.
Current regulatory practice for telephone service was
devised at a time when circuit-switched technology was dominant. Historically, regulators treated different types of networks
differently. Future regulation should, however, be based on a
91
Box 6.1: A VoIP Primer
“Voice over Internet Protocol” (VoIP) is a generic term referring to a technical standard that enables the transmission of
voice traffic, in whole or in part, over one or more networks that use Internet Protocol (IP).
Standards or “protocols” for VoIP are still evolving, but two main open protocols and proprietary vendor protocols enable
VoIP:
– “H. 323” – The most widely adopted protocol for the transmission of VoIP, this is an ITU legacy standard that builds on
earlier protocols for the transmission of voice and video over analogue PSTN, ISDN and ATM networks; and
– “Session Initiation Protocol” (SIP) – An application-layer control protocol, SIP is an end-to-end signalling protocol. SIP
facilitates communications between two or more SIP-supported devices, but it is not the only protocol required to make
VoIP calls, which take place via additional protocols.1
How VoIP technology works: Voice (or data) is compressed and converted into digital packets that travel over the internet
(or a private network utilizing VoIP) and are then converted back at the other end, correcting for echoes from end-to-end delay,
for jitter (variability) and for dropped packets. The data packets are non-isochronous and may take many different and independent paths to the intended destination, arriving out of sequence or with different end-to-end delays. VoIP technology makes much
more efficient use of bandwidth. Plus, voice is transmitted on IP-based networks at considerably lower cost than calls on circuitswitched networks, which require dedicated connections for the entire duration of the call.
VoIP Applications: The first generation of VoIP services, known as PC-to-PC or Class 3 services, allows individuals only
to call other people using the same service (examples include Yahoo! Instant Messenger). Voice signals transmitted are not routed
or switched through the PSTN at all. Second-generation VoIP services (PC-to-Phone or Class 2 services) allow calls from PCs
to any PSTN telephone number, including local, long distance, mobile, and international numbers (examples include Dialpad,
Net2Phone and Skype Out). A third generation of VoIP services (Phone-to-Phone, or Class 1 2 enables use of a traditional telephone to make VoIP calls, using an adaptor at the customer premises. The calls are then routed over an IP network rather than
the circuit-switched PSTN.3
1
See Webopedia at http://www.webopedia.com/TERM/I/Internet_telephony.html
2
Class 0 is phone-to-phone over the PSTN. This is David Clarke’s classification system (MIT).
3
A VoIP network requires a terminal or communication end-point, which can be a phone, PC or even a software programme. Terminals are identified by
at least one IP address (e.g. [email protected]) and are registered with a server, which stores IP addresses and can map an address to a terminal. The
server might also store location, identification and traffic data. Finally, gateways act as bridges between the local PSTN and IP network to allow calls between
different networks so that the signaling protocol can be understood between networks and so that IP addresses and regular PSTN numbers are recognizable
between networks. Signaling data is exchanged between switched circuit telephone networks and VoIP networks. This information is used to set up, manage
and release voice calls, and to support telephony services such as caller ID, toll-free calling, and mobile authentication and roaming services.
fundamental recognition of the convergence of telecommunications, broadcasting, media and information technology sectors. This means that all transmission networks and services
increasingly will be addressed by a single regulatory framework. For the remainder of this chapter, the term Information
and Communication Technologies (ICTs) will be used to reflect
this comprehensive, forward-looking approach.
For operators, VoIP represents three broad types of commercial opportunities: price arbitrage, savings from new network topologies, and new products and services. Each of these
is considered in the following subsections.
6.1.1.1
Price Arbitrage
Arbitrage is a term used to describe a situation where one
buys something at a cheaper price in one market in order to sell
it at a higher price in another. The growth of VoIP for international calling has been built on the wide gap between retail and
wholesale calling prices in many parts of both the developed
and developing world. These differences are a function of the
92
uneven introduction of competition in voice markets around
the globe.
In Africa, for example, it may cost a caller (at the retail
rate) between USD 0.50 (50 cents) and USD 1 a minute to
call Washington, DC, but the international operator may buy
the call (at the wholesale rate) for between USD 0.01-0.03 (13 cents). Where this circumstance exists, incumbent operators
are able to maintain high margins because they either have
monopolies or limited competition at the retail level. What
is often described as “bypass traffic” or the “grey market” is a
proxy for competition, particularly international voice calling.
This has already forced incumbent telephone companies to cut
their international rates. And because there is less regulatory
protection of international voice markets, these rates will continue to decrease.3
Moreover, many incumbent operators are going through
the process of “rebalancing” their tariffs in line with the costs
of providing services. In the pre-competition days, high international rates were used to cross-subsidize rates on domestic
networks. With competition in the largest voice markets driv-
C HAPTER 6
Figure 6.1: Shades of Grey
Grey Market Revenues as a Percentage of Overall International Call Revenues in Selected Regions and Countries
Africa
Brazil
Colombia
Costa Rica
* Source:
** Source:
*** Source:
**** Source:
20-30%*
36%**
50%***
20%****
Balancing Act
Abrafix
V-P Technology, Orbitel
Incumbent telco ICE
ing down international rates, cross-subsidization is unlikely to
be sustainable. Mandated tariff rebalancing will also end this
business model in many countries.4
ing, and some IP network components can even be purchased
in retail electronics outlets rather than as inflated “integrated
solutions.”
The VoIP service market has been fostered by the introduction of IP networks and the proliferation of Web-based
transactions. Grey market operators can simply sign up on a
website and gain access to international calling capacity that
they can resell to end users. Because these are data services,
the calls are not recorded as telephone minutes, and they need
not pass through the international gateway of the incumbent
operator. While quality-of-service issues do arise, many costconscious callers seem willing to make the trade-off between
price and quality.
Not surprisingly there is considerable debate about these
cost advantages. Some of the debate centres on reliability and
cost of the newer generation of network equipment – including
Wi-Fi and Wi-MAX – as part of an IP network roll-out. These
new wireless technologies can and are being deployed both to
create local loop VoIP access and for backbone links. Again, it
is argued that this is being done at prices that are much cheaper
than possible with traditional copper or fibre networks. As with
arguments about IP networks, the potential cost savings using
wireless technologies can be debated. But incumbents worldwide are deploying these technologies, at the same time they
are threatened by them.
As a commercial opportunity, however, much of this
market is entirely price-dependent. As international calling
prices plunge, the arbitrage effect may disappear. The legalization of VoIP services in a wider range of countries is, in effect,
the introduction of greater competition that will reduce this
price-arbitrage gap. The future of telephony revenue, particularly in the international domain, appears to be “low-margin,
high volume” rather than “high-margin, low-volume” for this
type of calling. Where VoIP has been legalized, VoIP providers
appear content to work with lower margins than established
operators.
6.1.1.2
Savings from New Network Topologies
Many of the world’s larger operators have been persuaded
to consider VoIP because an IP-based network can carry both
voice and data in one network rather than two. In this way,
operators can invest in a single network that can be used more
efficiently for different forms of traffic.
Moreover, IP network deployment costs often come in
smaller increments than those required for circuit-switched
facilities and dedicated circuits. It is possible to add capacity
incrementally, in a manner that will bring a return on investment more quickly than the traditional multi-million dollar
telecommunication infrastructure investments, which require
many years to produce the required return. For example,
operators can replace large numbers of traditional switches
with fewer “soft switches.” Smaller investments can often be
financed from cash flow rather than major external borrow-
C HAPTER 6
Similarly, mobile operators that have invested considerable sums in 3G licences and need to make a return on their
investment over 5-10 years are also threatened by the potential
of VoIP. They are particularly vulnerable to the effect substitution of mobile VoIP will have on operators’ high-priced international roaming services. This presents a recurring dilemma
for regulators. Should they protect the mobile operators’
investments and delay cost-saving innovations for consumers?
Or, should they allow wireless-enabled, mobile VoIP to flourish, potentially risking thwarting investment in 3G network
deployment?
These questions are particularly pertinent for regulators
and policy-makers in developing countries. There, the choice
is often between defending a government-owned incumbent
(for financial and social reasons) and making cheaper communications available to a wider number of people, particularly in
rural areas.
6.1.1.3
New Products and Services
The convergence of voice, data and video on IP networks
allows users to combine these different forms of traffic and
significantly expand the range of product and service offerings.
Many operators are now offering the so-called “triple-play”
option that combines all three in a single service. Senegal’s
Sonatel, for example, has rolled out a “triple-play” service
offering voice, -internet access and television programming.
93
This ability to package services together has implications
for competition, as users increasingly seek a single provider and
billing option. Triple play offerings have the potential to open
up television as a delivery platform for a far wider range of rich,
multimedia services. This could overcome, to some extent, the
lack of installed, internet-connected computers in developing
countries. This may be a solution for the urban poor, but it still
will not address lack of internet access in rural areas without
electricity or television coverage. Yet, even though the triple
play may not be everywhere overnight, it is certainly going to
be relevant in the medium-term.
In addition, there are now several VoIP crossover technologies coming into use on mobile phones. For example,
“push-to-talk” services are essentially what most people used
to understand as “walkie-talkies.” That is, they enable instantaneous, direct two-way conversations between two individuals, using IP software that rides on a mobile phone platform.
U.S.-based operator Nextel (based on its origins using a
trunked radio network) pioneered the push-to-talk experience, and others started offering the service in 2003. The product debuted in Europe with Orange’s “Talk Now” offering.
Although it requires users to subscribe to a data service from
one of the major carriers – and there are some service quality
issues – there has been significant take-up. Push-to-talk is sufficiently threatening the traditional “walkie-talkie” market that
Motorola has produced specially designed, rugged “push-totalk” phones at the beginning of 2005. Two of India’s mobile
providers – Hutchison Essar and Tata Indicom – also launched
“push-to-talk” services in May 2004.
Another area of development will be the current testing
of products that integrate cellular and WLAN networks and
provide voice from a WLAN device (See Chapter 3). Mobile
carriers in the United States and Europe are already rolling out
networks that offer 11 mbit/s 802.11 WLAN access. This could
eat into their existing data services (and perhaps even undercut
the rationale for 3G), but operators know they cannot ignore
the technology if competitors deploy it. One European UMTS
mobile operator has been sufficiently worried by the impact of
VoIP that it has threatened to block Skype calls to its subscribers.5
6.1.2 Changes in the Voice Business Model
These three categories of business categories – arbitrage,
new network topologies and new services – have come about
from a number of key changes in the underlying business
model for voice service.
6.1.2.1
The Impact of IP Network Features
With traditional telephony, “intelligence” in the network
is located centrally (in the functionalities of the switch) and
is usually controlled by one operator. Largely “dumb” devices
(telephones) are attached to the network, and they have a
limited set of functions. The traditional telephone network’s
root-and-branch structure means that traffic flows to and from
exchanges in ways that reinforce this pattern. For example,
traffic for international destinations is commonly routed via a
single international gateway. Telecommunication carriers main-
94
tain bilateral relationships with other carriers and exchange
revenue through the international settlements system.
By contrast, IP networks are not controlled by any single
entity, other than for the most basic transport to other networks. The “intelligence” is deliberately designed out of the
network architecture. Put simply, the network is “dumb” and
the intelligence is at the edge of the network, in the terminal
equipment. A computer accessing the network has a far more
complex range of service functionality in its application programmes than the network to which it is attached.6
IP traffic is routed via the easiest route, not through central choke points. International traffic can just as easily flow
from an ISP, a cybercafé or a telephone company. Each of these
has only to open a network connection and have the required
capacity available. In this way, VoIP traffic does not need to go
through an international gateway. The network design originated from United States military requirements for redundancy and resilience, and that led to a decentralized network
without a central focus or control point. And because of its
use for research, the internet was designed to be open to users
through publicly available standards, making it easy to access.
Due to their open nature, IP networks pose particular security challenges. The PSTN and mobile networks are
closed systems with controlled security and privacy. IP networks, however, have open architectures in which vulnerabilities, threats and communication security risks exist in various
network elements. Special measures are required for ensuring
communications security.
6.1.2.2
The Separation of Retail and Wholesale
Historically, a vertically-integrated organization like the
telephone company carried traffic and offered services, usually from a monopoly market position. In a more liberalized
market, the same telephone company might be selling international transmission to both external ISP customers and to its
own ISP. This can lead to accusations of conflicts of interest.
For VoIP service providers, the terms under which they can
access broadband networks become a key issue.
With liberalization bringing new market entrants, there
has been increasing discussion of separating the “retail” functions of the operator from the “wholesale” ones. In other words,
there is more talk of service provision and network operation as
separate roles. Alternative infrastructure providers such as utility companies have begun to reshape themselves into wholesale operators by adding bandwidth to their private networks.
Meanwhile, ISPs, VoIP service providers and “mobile virtual
network operators” or MVNOs (essentially, mobile service
resellers) have begun to retail services to end users. The nature
of IP networks has encouraged changes in thinking about these
two functions.
As a result, many telephone companies have internally
separated out their wholesale and retail functions in order to
better understand the underlying cost structures of different parts of their business. In some instances, this has been
prompted by regulators seeking to assess the costs and clarify
the terms for access to local loops or other network elements.
C HAPTER 6
In other cases, the companies themselves have wanted to analyse their costs more closely.
6.1.2.3
Changing Charging Structures
Wholesale traffic rates are largely the same on routes
between the more competitive markets in the world. So it is
no surprise that several VoIP service operators offer the same
or broadly similar retail rates for calls between these countries.
For example, Skype sells “Skype Out” minutes to enable subscribers to call PSTN phones in other countries. At the time of
publication, it cost less than USD 0.02 (2 cents) a minute on
Skype to call Australia, Chile, Europe and North America.
Most telephone calls have traditionally been billed according to where they originated and terminated. Pricing was
distance-sensitive, even where the actual cost of terminating
the calls was not. This model, however, has been turned on
its head by VoIP services. Now, per-minute charges are based
on what the market will bear in the terminating country, and
those countries with less competition are more expensive to
call. But even those countries can be called more cheaply using
VoIP than routing through their circuit-switched incumbents’
services.
As discussed in greater detail in Section 6.4.3, numbering has traditionally reflected geographic location, but with
VoIP, this is no longer the case. Many VoIP service providers
offer users “virtual” numbers, allowing them to be reached at a
“local” number when they are actually thousands of miles away.
For example, a VoIP customer may live in London, where her
PSTN dialling code is “207.” Her mother may live in Florida,
where she has a “561” PSTN dialling code. The VoIP service
provider can give its customer a “561” dialling code that rings
on her London line, allowing her mother to call her at the
VoIP rate and avoid international long distance charges. Argentina’s PVTEL, for example, offers its customers the choice of a
Buenos Aires or Miami dialling code.
Because fixed-line and mobile phones (rather than SIP or
“soft phones”) are still the dominant form of telephone terminal equipment, a hybrid service model is emerging in which
operators offer consumers a combination of broadband access,
plus an adaptor or VoIP-enabled phone. The package will also
include free calling to the company’s other subscribers (as an
incentive to recruit new subscribers by “word of mouth” sales)
and cheaper domestic and international calls.
6.2 The Pace of VoIP Market Development
All current market indications show that IP networks
and services eventually will replace PSTN networks and services and will alter the mobile business, as well. Moreover, the
introduction of IP networks will affect all countries, although
the timeline of adoption will vary widely. Some major international carriers have committed themselves to making the transition to VoIP, including:
C HAPTER 6
•
•
•
•
British Telecom (100 per cent conversion planned by
2009);
MCI (100 per cent of all traffic by the end of 2005);
AT&T (100 per cent by the end 2010); and
Telecom Italia (80 percent of all traffic was VoIP by the
end of 2003).
In Europe, the number of market players offering VoIP
increased with astonishing speed during late 2004 and early
2005. According to rough estimates, in March 2005 there were
at least 10 VoIP service providers in most Western European
countries, and in some the number exceeded 40. This kind of
growth appeared to be taking place across the globe. For example, there were 11 companies in Pakistan offering VoIP by early
2006, and more than 80 VoIP providers were licensed in Malaysia.7 Since 1 February 2005, all “value-added network” service
providers in South Africa were legally allowed to carry VoIP on
their networks. While there were no specifically licensed VoIP
providers, several companies began offering VoIP solutions
there in 2005.
In less liberalized markets, the impact of VoIP wholesale
transmission indicated how the market was changing. In 2004,
up to a quarter of all operators in Africa were using VoIP to
carry at least part of their international traffic (these agreements were politically sensitive, so establishing exact numbers
was difficult). In late 2005, Telkom Kenya was preparing to
offer a VoIP-based international service. Four African carriers
– BTC (Botswana), Mundo Startel (Angola), Telecom Namibia
and UTL (Uganda) – announced plans to introduce IP-based
networks, and the second national operator in South Africa
was expected to, as well.
Mexican incumbent Telmex had already implemented IP
for the majority of its core network, and various Mexican carriers (Alestra, Avantel, Axtel and Protel) have been conducting
initial trials pending legislative changes. Marcatel was already
offering IP-based long distance services. Oman’s incumbent,
Omantel, meanwhile, had committed to creating an end-toend IP communications services network.
The transition to VoIP is so rapid and far-reaching that
it is hard to make definitive statements about its progress. In
some countries, legalized VoIP operators are already offering
significant cost and service choices for both national and international calls. In others, the process of liberalization has not yet
begun, and the only “choice” for consumers is to look to grey
market operators. Irrespective of national regulation, there has
been rapid growth in VoIP services over the internet. Carriers
such as Delta Three, Skype and Vonage have increased their
global subscriber bases rapidly over the last three years.
Although there is little market data on the progress of
this transition, it is useful to differentiate between the different types of VoIP service transitions that are occurring. At the
wholesale level, there is a well-developed market for the carriage of international traffic over IP networks. The calls actually originate from a PSTN phone as circuit-switched traffic
and are then converted to packet-switched data at gateway
switches. They are routed internationally over IP networks
to save money and increase efficiency, then converted back to
95
circuit-switched calls at the destination. Domestic calls may
be carried over an IP backbone, with the same conversions
happening at either end of the call. Such calls may happen in
parallel to traditional routing over circuit-switched (PSTN)
networks. Packet-switched routing is essentially seamless to
the consumer.
At a local level, VoIP is a much more challenging undertaking. Few end users have IP-enabled phones or software yet
on their PCs or mobile phones. Also, the implementation of
enterprise-level, IP-enabled PABXs that can handle both timedivision multiplexing (TDM) and IP in corporate markets
varies enormously from sector to sector and country to country.
Investment in customer premises equipment, both at a consumer level (by the individual user) and by companies (at an
enterprise level) will take time. An unpredictable mix of customer choice, service availability and regulatory policy in each
country will drive market take-up of IP-based services and
equipment. The cost of the equipment needed – particularly
at a local level – will only begin to come down once there is a
sufficient volume of buyers.
The question posed by these developments is whether the
transition to VoIP requires a revolution in regulatory thinking,
or whether an incremental approach is wiser. The best way to
consider this question is to differentiate between short-term
changes, which are largely evolutionary, and long-term changes,
which are more far-reaching. The transition from short-term
to long-term changes is summarized in Box 6.2. The changes
described as “long term” indicate that VoIP represents a major
disruptive force for all telecommunication service providers.
In other words, when VoIP becomes the primary transmission
mode for voice calls, radical changes will need to be made in
many regulatory models.
Clearly, regulators will need new knowledge and expertise.
Nearly all regulators and staff are steeped in circuit-switched
technology and services. In the future, regulators will need to
understand the new, layered IP networks, the service concepts
based upon them and the influence these will have on the
future shape of the market. And because IP networks and VoIP
services change rapidly, regulators are under greater pressure to
make swift decisions and decide now on a course of action.
6.3 Grappling with Change: Regulators’ Responses to
VoIP
Before looking at the detailed issues that VoIP raises for
regulators, it is worth looking at the overall policy and regulatory responses of different countries so far. Those responses
are extremely varied, so this section groups countries under a
series of headings corresponding to their approach to VoIP.
Some countries have adopted an incremental, evolutionary approach to VoIP regulatory issues, making modest
adjustments to their regulatory frameworks. For others, VoIP
represents a considerable threat to the established order (particularly the international arbitrage aspect) and remains illegal.
Some countries simply have not dealt comprehensively with
VoIP yet. In China for example, the basic service operators
(China Unicom, China Telecom and China TieTong) have
96
driven deployment of IP technology. But there is currently no
specific VoIP regulation, and VoIP has not been classified as
either a value-added network service or as a basic service. Basic
telecommunication licensees are allowed to offer VoIP services
and use IP technology in their core networks. China’s government, however, is considering banning the provision of VoIP
services by anyone other than licensed operators. Cybercafés
that offer long distance calls are deemed to be operating illegally, and basic service licensees are seeking ways to shut them
down.8 Currently, ISPs can only offer PC-to-PC VoIP services.
6.3.1 A Liberalized Policy Approach to VoIP
Various countries have legalized VoIP services at different levels. For example, all forms of VoIP service are legal in
Canada, the European Union, India and Korea (Rep.). The following paragraphs provide other specific examples:
European Union: The European Regulators Group (representing regulators from 27 European countries) has agreed
to a common statement on the regulatory approach to VoIP.
According to the Group, VoIP should be used to enable (for
the benefit of consumers) the greatest possible level of innovation and competitive entry in the market, while ensuring that
consumers are adequately protected. VoIP rights and obligations should be interpreted in accordance with the European
regulatory framework, including the policy goals and regulatory principles existing today. Consumers and service providers should have enough information to make informed choices
about services and service providers.
United States: The 1996 Telecommunications Act separated telecommunication services and information services and
defined them differently. The FCC has formalized a policy of
not imposing traditional telecommunication rules on internet
applications, which are considered information services. At
the beginning of 2006, the FCC was engaged in proceedings
to examine issues raised by IP-enabled services, including VoIP.
These proceedings were examining various social issues, such
as how to adjust contributions to the Universal Service Fund,
and the regulatory classification of services.9
Japan: VoIP is permitted and is subject to minimal regulation. The legal framework distinguishes three types of VoIP
services based on the quality of the service. Providers that do
not need numbers for their operation (that is, providers of PCto-PC communications) do not have to comply with quality
of-service (QoS) requirements. If the provider can ensure a
minimum QoS (in terms of end-to-end voice quality and endto-end voice delay), the provider can receive “050” – prefix
numbers. Only if the quality is equal to traditional telephony
can the provider use the same numbers as PSTN operators.
Tariffs and access charges for VoIP services are not regulated in Japan. Interconnection is required only if the VoIP
provider is a facility-based operator. VoIP providers have to pay
access charges to terminate calls on the PSTN. It is also worth
noting that Korea (Rep.) adopted a broadly similar approach to
VoIP in September 2004.
Canada: Following a public consultation process, the
Canadian Radio-television and Telecommunications Commis-
C HAPTER 6
Box 6.2: The VoIP Transition
Short to medium-term evolution (evolution from PSTN to IP networks)
Technical developments
– PSTN phone services and VoIP services exist in parallel.
– PSTN – IP network gateways are needed in most cases.
– Standard (E.164) numbers are (mainly) used; ENUM use of E.164 numbers increases.
– Terminals: Adapter + regular phone, IP-phone or a software-enabled phone.
Transition period for the market
– New, inter-modal competition develops, featuring cost-structure advantages, innovative services (in particular, nomadic use
of VoIP) and lower charging models.
– Voice traffic is shifting to IP-based traffic and revenues from traditional phone services decline.
Regulatory model
– Changes are required to the current regulatory regimes to account for long-term influences.
– Regulators should balance basic main objectives:
• To enable the development of new services.
• To ensure acceptable social and consumer protection.
Long-term change (towards all-IP networks)
Technical developments
– IP networks and VoIP services are prevalent.
– Subscribers and services are addressed mainly by different types of internet addresses.
– E.164 numbers are likely to prevail at least in the global context.
– New types of terminals (e.g. combined GSM/UMTS/WLAN phones) support VoIP at home and in WLAN coverage
areas.
– VoIP is normally one service inside a large service set.
Market structure and competition developments
– Integrated, innovative and personalized services dominate the market.
– Nomadic use is important, increasing the amount of cross-border traffic and services.
– Cost and revenue models of service providers have changed radically.
– There is separation between the transport network and the services delivered on top of that network.
Regulatory models
– New legal frameworks and regulatory models are needed.
sion (CRTC) published a decision in May 2005 that it would
only regulate VoIP service when it is provided and used as a
local telephone service.10 The CRTC reached its decision based
on a principle of “service neutrality,” meaning that subscribers could use VoIP service interchangeably with standard local
exchange services.11 VoIP providers were classified in the same
category as competitive local carriers.12 The CRTC’s decision
called for:
•
•
•
•
•
•
•
Registration of VoIP resellers,
•
•
Access to numbers and local number portability,
•
Access to directory listings,
•
Equal access to interexchange carriers,
•
Rules governing “win-back” marketing,
In line with its approach to retail internet services, the
CRTC will not regulate computer-to-computer (peer-to-peer)
VoIP services that reside solely on the internet.
•
Comprehensive assessment by VoIP operators of access
for the disabled,
Singapore: In June 2005, Singapore introduced a new policy
framework for “IP telephony,” addressing the growing trend
C HAPTER 6
Message relay service,
Privacy safeguards,
Tariff-filing requirements,
Contribution to the national service fund,
Non-dominant carrier regulation,
The development of IP interconnection interface guidelines, and
The regulation of VoIP in areas where local competition
is not permitted (areas served by small ILECs and the Far
North).
97
of using the internet and other IP-based networks to make
local and international voice calls. The Info-Communications
Development Authority (IDA) planned to issue licences and
numbers for IP telephony providers. IDA’s regulatory framework called for minimal regulatory obligations. For instance,
operators providing PC-to-PC telephony services were not
required to provide number portability, emergency service
access, directory services, or to conform to QoS levels. But
operators must provide clear information to their subscribers
regarding whether their offerings allow access to emergency
services or meet the minimum QoS levels. Also, facilitiesbased operators were allowed to provide certain numbers only
to users with valid Singapore addresses, safeguarding the integrity of Singapore’s national numbering plan.
South Africa: As of 1 February 2005, any holder of a
“value-added network service” (VANS) or “enhanced service”
licence was allowed to carry voice traffic on its network. Before
that date, VANS providers were prohibited from carrying voice
traffic. VANS providers can now apply for numbers, spectrum
rights and interconnection with any operator. Various ISPs and
VANS operators have begun to offer retail VoIP, with aggressive
advertising. There is no rate regulation of VoIP services directly,
but the regulator is considering QoS issues and access to emergency services. It is worth noting that the planned second network operator’s service will be entirely IP-based, because of its
deployment of a fibre network.
The Philippines: In August 2005, the National Telecommunications Commission (NTC) issued new regulations treating
VoIP as a value-added service, for which only registration, not
authorization, is required. Commercial VoIP providers with
no network of their own are required to enter into interconnection agreements with network operators. Although such
interconnection agreements are to be negotiated between the
parties, the NTC will intervene where necessary to ensure
interconnection is provided under fair terms. Carriers that
have previously received authorizations were not required to
register with the NTC when starting VoIP services.13
6.3.2 An Incremental Approach to VoIP
Meanwhile, other countries have taken a more cautious
approach, making incremental changes rather than introducing
comprehensive new frameworks.
India: VoIP has been legal since April 2002, under the
designation of “Internet telephony,” which covers (1) PC-toPC voice (both within the country as well as abroad), (2) PCto-phone, and (3) “IP based H.323/SIP terminals.” Internet
telephony through PCs or IP-based terminals is available also
through India’s public “tele-info” centres and internet kiosks.
Facility-based operators can provide internet telephony and
use VoIP technology to manage their networks. Furthermore,
the Telecom Regulatory Authority of India (TRAI) has issued
regulations on quality for VoIP international long distance
calls, differentiating between two quality levels: toll quality and
below-toll quality. TRAI does not regulate tariffs of VoIP services offered by ISPs.
Bolivia: VoIP is considered telephony. In January 2005, a
Bolivian ISP, Unete, announced an investment of USD 5 mil-
98
lion to launch a national and international long distance voice
service.
Ecuador: VoIP providers are required either to have a
licence for local or long distance public telephony or to establish resale agreements with licensed operators. In February
2005, the regulator CONATEL published regulations covering
cybercafes and tele-centres. The regulations limit the number
of PCs that can be used for VoIP services to no more than 25
per cent of the total – or one PC if the cybercafé has only two
or three computers.
Honduras: The regulator has allowed VoIP services, provided that operators contract with the monopoly incumbent,
Hondutel. The organizations that do this are described as “suboperators” and they can use their own networks to sell other
licensed services.
6.3.3 VoIP Consultations
Many countries have not resolved how they will regulate
VoIP and are engaged in public consultation proceedings.
Chile: In July 2004 the Chilean regulator, SUBTEL,
launched a public consultation, indicating that all services
offered through some part of the PSTN network should be
under PSTN regulations. But services provided entirely over
the internet would not be subject to the same conditions. The
regulator suggested creating a broadband voice licence that
would cover VoIP. Some operators responded that the proposed
classification was too rigid and potentially problematic in an
increasingly converged environment. The Chilean incumbent
argued that the introduction of VoIP would rob it of income
while benefiting only a small group of the population.
Colombia: In June 2004, the Ministry of Communications
issued a consultation document on VoIP services. The consultation was completed, but no action had been taken by August
2005. Prior to any results of the consultation, rules required
operators to obtain a basic PSTN-service licence, but the use
of a PC to make calls over the internet was not restricted. The
Ministry was seeking to include VoIP in the existing public
telephony category, but several parties in the consultation suggested that VoIP merited a new service category.
Jordan: In May 2005, the Telecommunications Regulatory
Authority (TRA) of Jordan issued a consultation document on
the delivery of IP voice services. This raised several issues for
comment, including (1) whether or how to distinguish among
different types of voice services that are identical to the consumer; (2) the provision of information to equipment purchasers and potential users; (3) the roles of network operators and
service providers in maintaining network integrity; (4) geographic and non-geographic numbering; (5) emergency service
requirements; (6) interconnection; (7) using class licensing for
VoIP service providers; and (8) quality of service issues.
Hong Kong, China: In June 2005, Hong Kong’s Office of
the Telecommunications Authority (OFTA) published a statement on the “Regulation of Internet Protocol (IP) Telephony.”14
This statement outlined OFTA’s position following a consultation process begun in October 2004. OFTA said service-based
providers should be allowed to compete with facility-based
C HAPTER 6
operators. Moreover, it upheld the principle of technological
neutrality. Therefore, OFTA decided to introduce two different licenses for VoIP providers:
•
Class 1 services – IP telephony offered with service attributes
similar to those of conventional telephone services; and
•
Class 2 services – Those that do not have the same attributes
as conventional telephony.
Class 2 services were subjected to minimal regulation,
although service providers were required to inform customers
about the limitations of their services. Class 1 service providers would have to fulfil basic telephone licensing conditions.
Meanwhile, Algeria, Israel, Taiwan, China and Trinidad and
Tobago have also begun consultation proceedings on VoIP.
Kenya has issued guidelines legalizing various categories of
VoIP, following public consultation.
6.3.4 Where VoIP is Illegal
VoIP remains illegal in quite a few countries.15 Where it
is, governments adopt a variety of strategies to eliminate grey
market operators. Some regulators seek to ban websites that
allow users to make international calls. Others periodically
confiscate the equipment of grey market operators. Some jurisdictions back up these sanctions with jail sentences.
Before the end of Panama’s telephone service monopoly
in 2003, the Public Services Regulator required all ISPs to
block IP ports identified with VoIP services. In addition, sometimes telephone companies filter (block) VoIP services on their
own. For example, one ISP in Mexico filters out VoIP service
providers including Skype. Operators in Kenya have also filtered VoIP traffic. In almost all of the countries where incumbents provide estimates of the grey market, the amount of “lost”
traffic indicates that few of these blocking strategies work completely. Usually, enterprising consumers and companies are
able to access some form of VoIP service.
6.3.5 Classifying VoIP Services for Regulation
As can be seen from examining the current state of VoIP
regulation, a key question for regulators is how to classify various forms of VoIP, IP telephony or “Internet telephony.” Any
rigid classification set at this juncture is unlikely to be stable,
given the pace of technological and market-driven change. Any
classification used also depends on national policies and legislation. Box 6.3, however, offers three general categories.
Below are various examples of how different countries
have treated classification of VoIP services:
The European Union: The regulatory framework addresses
the question of how communication services should be classified in two ways. The Universal Service Directive classifies
services for the purpose of consumer and social protection. For
market and competition management, the need for regulation
is assessed by analysing whether VoIP services have “significant
market power” (known in some other countries as “dominance”) in one or more “relevant markets.” The services covered by the Universal Service Directive are divided into two
categories:
C HAPTER 6
•
Electronic Communication Service (ECS) – includes
services provided for remuneration and consisting wholly
or mainly in the conveyance of signals on Electronic
Communications Networks. ECS is treated with lighter
regulation.
•
Publicly Available Telephone Service (PATS) – consists of a service that includes all the following functions:
available to the public; used for originating and receiving
national and international calls; and provides access to
emergency services through a number or numbers in a
national or international telephone numbering plan. PATS
attracts more regulation and obligations. The main obligation compared with ECS is the provision of emergency
calls.
The EU regulatory framework seeks to be technologyneutral, although some of its rules are based on traditional
telephone technologies. Debate on the classification of VoIP
services is ongoing. Meanwhile, different European governments seem to be interpreting the regulatory framework for
VoIP services in divergent ways. These vary from a flexible
reading of the rules to a strict interpretation of the framework’s
wording. The following approaches are evident:
•
One approach, apparent in the United Kingdom, is a flexible classification regime, in which service providers, not
regulators, decide in which regulatory category they want
to be classified;
•
Another approach distinguishes different types of services
based on a strict interpretation of the PATS definition.
That is, PATS obligations are imposed on a voice telephony service only if all four parts of the PATS definition are
fulfilled;
•
A compromise approach is to classify a service as PATS if
it is available to the public for originating and/or receiving national or international calls through a standard
telephone number. Access to emergency services is not
regarded as being an integral part of the PATS definition,
and whether to mandate such access can be decided separately.
For competition policy purposes, each EU national regulator assesses how different VoIP services fit into the EU’s
“relevant markets” categories. At present, there is very limited
practical experience and precedent for VoIP, and the EU is currently discussing market analyses that include VoIP services.
Canada: The classification of VoIP services stems from
discussion of emergency call services. From this perspective,
there are currently three different types of VoIP service offered
to customers: fixed, nomadic and foreign exchange. Users of
fixed VoIP service can only place a telephone call from the
location where the service is provided. Users of nomadic
VoIP service can make calls from any location where users can
access their service via the internet. Foreign exchange VoIP
service allows users in one exchange to receive telephone calls
dialled as local calls in another exchange that they have selected
(for example, a customer can be reached in Ottawa by calling a
Halifax local telephone number). These different types of VoIP
services have different obligations regarding emergency calls.
99
Box 6.3: Classification of VoIP Services
Regulators around the world face the challenge of trying to differentiate between types of VoIP services and capabilities. One
approach to classification of VoIP follows:
Category I
VoIP offerings that do not require regulation because there is no clear service provider. This would cover PC-based VoIP communications installed by end users via software downloads (Examples include GIZMO, Yahoo Instant Messenger and Skype).
Category II
VoIP offerings that are outside the scope of regulation because they are not directly offered to end users or do not impact the
PSTN, including:
– Corporate private networks, where VoIP is used to provide internal communications,
– IP technologies used within a public operator’s core network, but which do not affect the retail services offered to end users.
Category III
This category would cover publicly available services provided to end users. There are many different kinds of publicly available VoIP service offerings, and the regulatory treatment depends on the nature of the service being offered and relevant national
legislation.
A large number of national regulatory agencies have carried out consultations on VoIP issues but have not reached any final
classification decisions. Many are trying to adapt service classifications provided under existing telecommunications legislation
for VoIP services. There appears to be a consensus, however, that VoIP services residing solely on the internet (PC-to-PC calling)
should not be regulated. Global discussion of these issues focuses on approaches to VoIP services that are similar to those described
in category III above. The basic regulatory question that hovers over the whole discussion is whether or not VoIP can be regarded
as a substitute for PSTN telephony.
United States: There has been substantial debate about how
to classify (and therefore regulate) VoIP services. The country’s
Communications Act (as amended by the 1996 Telecommunications Act) applies more stringent regulation on “telecommunications services” services than on “information services,”
and there traditionally has been reticence to overly regulate
internet-related services. From the FCC’s perspective, services
that are only provided over the internet (like MSN Messenger and Skype) are classified as information services, and even
where VoIP services have a gateway to exchange traffic with the
PSTN, they should only be regulated in relation to emergency
calls and lawful interception.
6.4 Crafting New Regulatory Approaches to VoIP
6.4.1 Balancing Different Policy Needs
The transition to IP networks and VoIP services tends to
produce conflicting policy approaches in different countries.
This section defines some of these conflicts, which tend to
reflect opposing policy goals and sector objectives.
The main challenge is to balance short- and long-term
policy and regulatory approaches. In some countries, VoIP is
seen as a major threat to established operators because it undercuts their domestic and international long distance rates and
radically reduces their revenues. But strict regulation of VoIP
in the short-term to protect the incumbent’s revenues might
harm the long-term development of the sector. Often, this
drive to protect the incumbent stems from a particular social
policy, such as extending universal access. But regulators need
100
to consider not just the established operator’s point of view, but
also the welfare of end users and potential new market entrants.
The impact of lower prices from competition brought about
by VoIP will directly benefit consumers. Competition will also
help to increase the number of overall subscribers and boost
the volume of usage on all networks. In these circumstances,
it is important that policy decisions be based on trend data and
that the regulator analyse different market scenarios.
Regulation should promote competition, but it also needs
to protect consumers and take social concerns into account,
including universal access objectives. Regulators may find
themselves torn between the need to react quickly to new concerns and stepping back to see the shape and dynamics of the
emerging market. Many regulators are guided by a principle of
limiting regulatory obligations so as not to discourage market
entry or the introduction of innovative new services. This
leads to efforts to implement “light-touch” regulation in order
to encourage market and technological experimentation. For
example, some developing-country regulators have encouraged
the use of free or low-cost spectrum, or reduced licence fees,
in areas where there are little or no voice or data services.
A vital task for regulators and policy-makers is to manage
the transition to the new world of IP networks. The most basic
questions to address include:
•
How long should a PSTN network be maintained and
supported?
•
How much time is needed to make changes to existing
legislation and regulations, and how can legal stability be
maintained in a time of technological flux?
C HAPTER 6
•
How quickly should competition policy change to reflect
convergence on IP-based networks?
not yet finalized a policy on numbering and spectrum access
for VANS.
Governments do not have the luxury of setting the pace
at which they will resolve these questions. Market changes are
already under way, and they are likely to proceed whether they
are legally sanctioned or manifested in grey market activities.
Unfortunately, there is no consensus on how to address regulatory change prompted by the emergence of VoIP, either at a
global or a national level. The following subsections examine a
number of potential responses in key issue areas.
Other countries restrict market entry of new VoIP operators in various ways. In many countries, ease of market entry
depends on whether VoIP is defined as a voice or information/data service. Where VoIP has not yet been categorized, the
issue is often under extended debate.
6.4.2 Market Entry
In countries where telecommunication competition is
already legal, regulators will have to manage the range of issues
associated with the transition of services to IP networks. In
those countries where the PSTN remains an exclusive preserve of the incumbent, regulators face a more complex task.
These regulators have to manage two simultaneous, parallel
processes: (1) the transition from a monopoly to a liberalized
market, and (2) the transition from a circuit-switched PSTN
to an IP-based environment.
In the past, regulators largely dictated market changes, and
did so at a measured pace. In the future, however, regulators
will face a rapidly changing market based on a new type of
infrastructure. They will need to deal with a large number of
new services entering the market that may never have existed
before. The policy and regulatory framework may be influenced or changed by international market developments that
regulators cannot control or even influence. So it is becoming increasingly difficult to plan or orchestrate market developments in advance. There are widely differing strategies for
addressing market entry.
The European Union: Here, the approach is to facilitate easy
market entry by requiring individual licences only for scarce
resources like spectrum. For other network or service developments, only a notification or registration is required. The
EU also adopts a technology-neutral approach, leaving market
players to decide themselves what technology to deploy.
South Africa: The government has created conditions for
easier market entry in Value-Added Network Service (VANS)
or enhanced services, but it retains control over the market
structure for fixed, mobile and satellite services. Under the
current Telecommunications Act, only VANS and “Private
Telecommunications Network Service” (PTNS) licences may
be issued on a non-exclusive basis. Fixed-line service has been
dominated by the incumbent, Telkom, with a second entrant
set to commence operations in 2006. Mobile voice and data
services are supplied by three operators.
Until 1 February 2005, Telkom had a full monopoly
on facilities provisioning, and the resale of spare capacity for
VANS and PTNS was prohibited. Since then, VANS licensees
may obtain alternative facilities from other operators. New
regulations for VANS gave providers the right to access spectrum, apply for numbering allocations and interconnect with
other operators. Numerous VoIP providers are emerging in
the South African market as a result, although the regulator has
C HAPTER 6
Brazil: The regulator, ANATEL, has not defined VoIP as a
telecommunication service, a value-added-service or a technology. If VoIP is considered a data service, operators will need a
licence for multimedia communication services. And in order
to initiate and terminate calls outside of a private network,
operators need a licence for public switched fixed telephony,
which entails certain requirements for coverage and QoS.
Moreover, the licence application process is more complex.
Some countries, such as Colombia, Egypt and Nigeria, are
pragmatic about certain types of VoIP such as like PC-to-PC
telephony, which is regarded as a use of personal computing
equipment that would be almost impossible to control. Other
countries, such as Guinea, have legalized the use of VoIP over
virtual private networks (VPNs), which, again, is hard to detect
and control. In Nigeria, the regulator has stated that VoIP is
legal, provided operators obtain the appropriate licences. As
a result, one satellite operator with an international licence is
offering a VoIP service to its customers. Yet, other countries
completely restrict any form of VoIP market entry and seek to
control grey market operators through a variety of strategies.
6.4.3 Numbering Resources
Telephone numbers can be used for several different purposes. They can be used to differentiate between services and
inform users of tariff categories like premium call services.
Numbers can also be used as a tool to control markets by setting restrictions on the use of certain numbers. So access to
numbers – or more correctly, withholding of access to them
– can become a barrier to market entry.
VoIP calls can be routed to end users in several different
ways: IP addresses, SIP addresses, H.323 addresses or E.164
numbers. Traditionally, E.164 numbers have been needed
to originate and receive voice calls, but they may lose their
dominant position in the future and become just one of many
options. E.164 numbering ranges are usually divided into several generic types, indicating the services that may be offered
using them. Geographic numbers or special number series are
regarded as most relevant for VoIP services. Also, mobile, personal and corporate numbers can be used to address VoIP subscribers. These are, however, seen as less attractive in many
countries because users will associate them with high retail
calling prices.
The current position on the availability of geographic
numbers for VoIP services varies among countries. The main
argument in favour of allocating geographic numbers to VoIP
services is that they offer the best support for competition,
especially when combined with number portability. The main
arguments against this approach have been the nomadic nature
of VoIP and potential for exhausting geographic numbering
101
resources. There are three ways to allocate geographic numbers in order to support VoIP services:
• Allowing nomadic access in a limited area;
• Allowing nomadic access countrywide but requiring some
relationship with the geographic area of the number; or
• Removing any requirement for a relationship to a geographic location.
Regulators may also open new number ranges for nomadic
VoIP services. Broadly speaking, there are three types of possible new number ranges: (1) a general-purpose number range;
(2) a number range for nomadic services, and (3) a number
range for Electronic Number Mapping (ENUM)-based or
similar, software-based services.
ENUM is a protocol developed by the Internet Engineering Taskforce (IETF) that defines a domain name system
(DNS)- based architecture and protocol aimed at using a telephone number to look up a list of IP service addresses (e-mail,
IP phone addresses, URL, SMS, etc). The idea of ENUM is to
use an E.164 number as the key to identify the available communication services to contact a person. For VoIP, it is used
to route a VoIP call to an IP network based on the receiver’s
E.164 number.
The idea of creating new number ranges to accommodate
VoIP seems to have mainly been motivated by a number of
factors, including avoiding the impression that these are hightariff numbers like those for mobile phones. Also, regulators
may want to keep existing number ranges intact and avoid
having them depleted.
In Europe for example, the number ranges open for VoIP
use vary due to different numbering policies and the regulations related to them. The geographic number ranges are open
for VoIP services in most countries. But some countries list a
number of requirements that have to be fulfilled in order for
VoIP providers to obtain numbers.
Meanwhile, the cost of numbers (national numbering
fees) can be a significant barrier to market entry in some countries. Geographical numbers are typically allocated in blocks
(normally blocks of 1000 or 10,000). Where geographic numbers are sold with limited nomadic capability, VoIP service providers may need to get number blocks that cover the whole
country. The costs of doing so can be high enough to constitute an entry barrier for small providers. In Europe, numbering
costs vary greatly.
Number Portability is a key enabler of competition
because it allows users to retain their telephone numbers when
they change service providers. Regulators must decide whether
VoIP providers are included in number portability systems,
making them co-equal with all other competitors.
6.4.4
Emergency Calling
The nomadic capability of VoIP services poses a problem
for the provision of emergency calls because it breaks the link
between the calling party and the location. In the PSTN environment, calls routed to the emergency response centre carry
information on the location of the caller. With VoIP calls, however, there is – at least currently – no automatic way to convey
102
location information about the user calling an emergency service. Technical solutions, however, are expected to emerge from
the current standardization work that is being undertaken.16
The problems of handling emergency calls from VoIP
users can be divided into two categories: (a) emergency calls
made within a country and (b) cross-border emergency calls.
The first category is likely to be less problematic, due to the
likelihood of increased cooperation between service providers
covered by national regulation in a single country.
Cross border VoIP emergency calls raise more difficult
issues, because an emergency call needs to be identified as
such. Globally, there are more than 60 national emergency call
numbers (for example, “911” in the United States and “112” in
the EU). For an emergency call to be routed to the right country and the correct emergency centre, intensive international
agreement, cooperation and arrangements will be required.
Cross-border emergency calling is mostly an issue, of course,
in places such as Europe, which are interlaced with multiple national borders. In North America, by contrast, the vast
majority of emergency calling is entirely domestic, and crossborder arrangements are less of an issue.
From a consumer point of view, the best possible situation is to be able to reach emergency services from any handset, including a VoIP-enabled phone. Technical solutions will
be required to overcome any limitations in routing of emergency calls in a VoIP environment. A key question is what will
happen with VoIP connections during electricity outages, especially where VoIP is the only option for making a call in an
emergency situation. Various principles may be considered to
address this issue:
• If VoIP handsets or terminals are likely to be used for
making emergency calls, they should be equipped to do
so;
• An emergency call from a VoIP handset or terminal should
reach an emergency centre in the country in which the
call originates;
• Where possible, an emergency call from a VoIP handset or
terminal should reach the specific emergency call centre
that is responsible for receiving emergency calls for the
area in which the caller is located;
• The VoIP call made to an emergency centre should carry
Calling Line Identification (CLI), which can be used to
call back the person reporting the emergency if the person
is disconnected before full information has been provided;
• Where possible, the number provided by CLI for an emergency call from a VoIP handset or terminal should not be
linked to location information that is incorrect or misleading.
The following examples of VoIP emergency calling suggest various approaches:
European Union: The Universal Service Directive allows
for technical feasibility when imposing obligations relating to
the provision of location information. It states that location
information must be “handled in a manner best suited to the
national organization of emergency systems and within the
C HAPTER 6
Box 6.4: Emergency Services under the North American Numbering Plan
Both Canada and the United States are among those countries that follow a regional numbering scheme, known as the North
American Numbering Plan (NANP). Consequently, both countries have a standard emergency calling number: “911.” The existing local telephone networks currently provide two types of 911 services: Enhanced 911 and Basic 911. Enhanced 911 (sometimes
known as “E911”) service automatically sends customer location information to an emergency centre, where an operator dispatches
a response service. E911 allows dispatchers to send responders to the correct location even if the caller is incapacitated, agitated or
otherwise unable to provide an accurate description of the location. The older Basic 911 service, meanwhile, connects the caller to
a central call centre, which then connects the call to the correct emergency response centre. At that point, the caller must be able
to identify or accurately describe the location in order for responders to be dispatched to the right place.
Box 6.5: The FCC’s Enhanced 911 Service Order1
Provisions of the U.S. Federal Communications Commission’s May 2005 order on VoIP access to emergency services include:
• Interconnected VoIP providers must deliver all 911 calls to the customer’s local emergency operator. This must be a standard,
rather than an optional feature of the service.
• Interconnected VoIP providers must provide emergency operators with Calling Line Identification and location information
for their customers (an Enhanced 911 service) wherever the emergency operator is capable of receiving it. Although the
customer must provide the location information, the VoIP provider must provide the customer a means of updating this
information, whether he or she is at or away from home.
• Interconnected VoIP providers must inform their customers, both new and existing, of the Enhanced 911 capabilities and
limitations of their service.
• The incumbent local exchange carriers are required to provide access to their Enhanced 911 networks to any requesting telecommunication carrier. They must continue to provide access to trunks, selective routers, and E911 databases to competing
carriers. Interconnected VoIP providers must submit to the FCC a letter detailing their compliance no later than 120 days
after the effective date of+ the order.
1
Federal Communications Commission, 911 Services, at http://www.fcc.gov/911/enhanced/
technological possibilities of the networks.” Authorities should
“make caller location information available to authorities handling emergencies, to the extent technically feasible, for all calls
to the single European emergency call number 112.”
How the Directive is actually applied varies, depending
on national emergency systems and the capabilities of the networks involved. Similarly, related legal requirements – such
as providing caller location information, routing calls to an
appropriate emergency centre and providing CLI – vary greatly
between countries. Where VoIP services are regulated, some
countries have set the same legal requirements for both VoIP
and PSTN calls, although they provide a temporary technical reprieve from these requirements for VoIP calls. Currently,
nomadic VoIP service providers can only meet national legal
requirements for emergency services in a few countries.
Canada: CRTC decided in April 2005 to require providers of fixed (non-nomadic) VoIP service to provide the same
level of emergency service that incumbent telephone companies provide to their existing customers (either “Enhanced 911”
or “Basic 911” service). Implementation was made mandatory
within 90 days.17 CRTC required providers of nomadic and
C HAPTER 6
foreign exchange VoIP services to provide an interim solution
equivalent to basic 911 service. In addition, the Commission
required all VoIP providers to inform their customers about
any limitations of their offerings in providing access to 911
service. The VoIP service providers must secure the customer’s
express acknowledgement that they are aware of these limitations before starting service.
United States: The FCC issued an order in May 2005
requiring interconnected VoIP providers to provide Enhanced
911 Service (See Boxes 6.4 and 6.5). The order affects providers
of services that are functionally equivalent to circuit-switched
telephony. For example, the FCC included service provider
Vonage in this category, because its service enables customers
to send and receive calls from the PSTN. The order did not
impose E911 obligations on providers of other IP-based services, such as instant messaging or internet gaming. Although
these services may contain a voice component, the FCC said,
customers of these services cannot use them to exchange calls
with the PSTN. The FCC has also stated its intention to adopt,
in a future order, an advanced E911 solution that automatically
determines a customer’s location.
103
6.4.5 Universal Access and Universal Service
As customers switch to VoIP and disconnect from the
PSTN, regulators in many countries fear that a decline in
revenues from traditional carriers will undercut funding for
universal service funds, which often rely on mandated carrier
contributions. Of course, this begs the question of whether
VoIP providers should be included in universal service funding
arrangements. There are two issues at stake here: (1) potential
contributions to universal service funds by VoIP providers, and
(2) whether VoIP can be an effective tool to deliver cheaper
calling to a wider number of consumers.
United States: The FCC is confronting a multifaceted challenge to the financial health of the Universal Service Fund
(USF), stemming from lower wire-line telephony revenues.
The Commission is considering overhauling the contribution formula, which has been revenue-based. One suggestion
is to require contributions for any service that requires use of
a telephone number – which could take in some VoIP services. While the service providers would make the contributions,
they could pass the costs along to customers (as carriers now
commonly do) in retail fees or charges.18
Various other countries that allow VoIP, including the
Czech Republic, Mauritius, the Slovak Republic and Venezuela, levy
universal service/access contributions on operators. In Canada,
the CRTC has ruled that if the VoIP service provided allows
for access to and/or from the PSTN, the service is considered
eligible for contribution requirements, even if the customer
uses the service to make peer-to-peer (PC-to-PC) calls.
Regulators have to consider the potential consequences
of their decisions on funding of universal service programs. If
they continue to levy contributions on a narrow base of incumbent or traditional telephony revenues, they risk overtaxing a
declining base as subscribers flee to VoIP. But if they broaden
the base to include VoIP providers, they may raise the bar for
market entry by potential voice service competitors.
Indeed, IP-based transmission is arguably less costly and
more efficient than circuit-switched transmission. Using the
same bandwidth, a VoIP network can carry many times the
number of voice calls as a circuit-switched network, making
the transport cost, per bit of information, lower on packetswitched networks. For that reason, VoIP can be embraced as
a new tool in achieving universal service and universal access,
at least from an affordability standpoint. The Association of
Infocentres of El Salvador (Infotel), for example, is launching
a VoIP service for international calls using pre-paid cards. The
service will be available in 41 of the Association’s centres. The
initiative is supported by the Salvadoran regulator as a means of
reducing international calling costs. Similarly, the state-owned
Telecommunications Office (Telof) in the Philippines plans to
launch VoIP services in un-served rural areas.19
6.4.6 Competition
Consideration of how to regulate the VoIP market raises
the question of what kind of competition issues might arise
specifically in relation to VoIP. There are several. For example,
the development of VoIP service depends greatly on the avail-
104
ability of broadband access. Regulators have to ensure open,
non-discriminatory and fair-priced access for ISPs wanting to
resell broadband access.
Another challenge is to prevent incumbent operators from
blocking access to VoIP services by closing ports used by VoIP
services or refusing to lease facilities. There is no consensus at
this juncture about which regulatory tools should be used to
ensure fair competition. But it is important for regulators to
prevent anti-competitive behaviour against VoIP providers as
they apply downward pressure on tariffs.
European Union: Market regulation is based on the circumstance in each market. The telecommunication sector is
divided into 18 “relevant markets,”20 which can be analysed by
the national regulator in each member country. If robust competition is present, no regulation is applied, but if the regulator finds an operator with “significant market power,” it will
implement regulatory remedies to prevent that operator from
abusing its market power.
Several European countries are in process of considering
whether VoIP should be included in the relevant market covering fixed telephony. Again, results vary considerably, but in
some countries, VoIP services with a gateway to the PSTN are
considered part of the fixed telephony market for the purposes
of competition analysis. The whole question of where VoIP
services will be placed in terms of “relevant markets” will soon
be considered by the European Commission.
Canada: CRTC also has decided that VoIP should be seen
as part of fixed telephony market. But it will only regulate VoIP
services when they are provided and used as a local telephone
service. This decision was aimed at building sustainable competition in the local telephone market. Incumbent local loop
carriers (which are held to possess market power) are prohibited from pricing their local services below cost as a way to
stifle competition.
6.4.7 Interconnection
6.4.7.1
VoIP’s Challenge to Current Models
VoIP services challenge current interconnection models
in several respects, both from an economic and structural perspective. Again, it is helpful to separate what may happen in the
short-term, during the transition period to IP networks, and in
the longer-term, when most networks have become IP-based.
In the short-term, interconnection will be mainly between
IP networks and the PSTN. But over the long term, interconnection will increasingly occur directly between IP networks.
Short-term interconnection issues: Current interconnection
models are increasingly based on cost-oriented charges. VoIP
may disrupt this because of the difference in investment scale
between IP-based networks and PSTN networks. Although
views differ as to the exact scale of difference, the cost of IP
networks is significantly lower. So there is a possibility of variations in call-termination rates based on cost differences. There
are not likely to be any issues when the call comes from an
IP-based network to the PSTN, because the receiving PSTN
operators normally charge the same termination fee regardless of which network the call comes from. But when the call
C HAPTER 6
originates on the PSTN and terminates on an IP network, the
termination cost is difficult to determine. The relevant elements to assess the actual costs are unclear. Thus PSTN-to-IP
interconnection is likely to generate a difficult debate between
the different players, requiring regulatory oversight and intervention.
Long-term interconnection issues: In the long term, when IPto-IP interconnection is predominant, the application of current telephony interconnection models will create a number
of problem areas. These are examined below, and it should be
noted that these issues apply to all IP services and not just to
voice calls.
• Support of new IP-based services: IP-based networks
are expected to support new services, including thirdparty services. The existing, usage-based format for interconnection would call for interconnection agreements
and charging arrangements with each service provider, at
each interconnection point. But there is a basic technical
problem: there is no way (at least in the near future) to
transmit the charging information between IP networks.
These kinds of practical issues could constitute a barrier
to the roll-out of new services if changes are not made in
the interconnection model.
• Changes to cost structures: Developments in technology and huge economies of scale have resulted in the
substantial decline of core or backbone network costs.
The existing regulatory and commercial models assume
an expensive backbone network, which is why there is
so much emphasis on competition and pre-selection for
long distance and international calls. With less-expensive
IP backbones, there is a need to consider adopting capacity-based charging instead of usage-based prices for basic
connectivity. A simple approach would be to make one of
two changes: either separate charges for services and connectivity, or adopt “bill and keep” interconnection.
• Changes to the retail market: The existing retail market
is changing. Call prices are dropping, and some operators are starting to offer flat-rated tariffs with unlimited
call volumes for a fixed subscription. This sets up the risk
of arbitrage. Operators would benefit from having interconnection arrangements that better match the structure
of the retail charges. This change is causing many commentators to say that the days of time-based call charges
are rapidly disappearing. But there are still a number of
high-priced, time-based charging structures (for mobile
calls and premium-rate services), and providers of these
services will likely seek to continue these revenue streams
or find substitutes, such as metered charges for video or
music downloads.
6.4.7.2
New Approaches to Interconnection
These developments suggest the need to look for new
approaches to interconnection. A variety of ideas are under
discussion, and there is a lively debate between the “telecom
world” and the “internet world” on two basic approaches:
a) An open, internet-type approach, in which the separation
of service provision and connectivity occurs as it does on
C HAPTER 6
the internet. This would require separate consideration
of:
• Interconnection (interoperability) at the service level,
where services are charged on the “bill and keep” principle (peer to peer), and
• Interconnection at the connectivity level, where charging between networks is based on capacity charging or
another similar method;
b) A next-generation network (NGN) architecture approach,
in which network operators have more control over, for
example, service quality (such as providing different categories of guaranteed bandwidth), security (such as customer ID, authentication and security tunnelling) and
charging for services by third parties. NGN architecture
includes additional software, which is not present on the
basic internet network, such as IP Multimedia Subsystem
(IPS), which controls the interconnection of services to
networks. This means that in NGN architecture – since
network and application services are separate – network
operators can get a share of revenues from application
services.
In the future, there will likely be more than one interconnection model. Market players may be able to choose the one
that best fits their needs and situations. During the transition
period to IP networks, new interconnection models will perhaps exist in parallel with older arrangements, just as the internet’s current charging model works alongside the telephony
charging structure.
6.5 End User and Consumer Issues
Because VoIP services ride on an entirely new network
and transport model, there are inevitably issues with regard to
duplicating or recreating some of the salient features of the old
environment. These issues relate to quality of service (QoS),
network security, privacy and access for legitimate law enforcement purposes. This section examines how regulators and
other government authorities are addressing these issues.
6.5.1 Quality of Service
One of the requirements for the deployment of VoIP networks is the ability to offer toll-quality service equivalent to
the existing PSTN. The QoS standard for VoIP can be defined
in several ways, depending on whether it is considered from an
end user or a technical perspective. The end users’ perceptions
of service quality can be measured through subjective quality
assessment. The most common consumer-based method to
quantify QoS is the “Mean Opinion Score” (MOS) developed
by ITU-T.
6.5.1.1
End-to-End Quality
Discussions of VoIP QoS typically highlight the issue of
increased end-to-end delay and discuss the effects of this delay
in interfering with the normal cadence of voice conversations.
People expect their conversations to be in real time. If delays
occur, people begin to talk over each other and conversations
begin to break down.
105
This problem is actually a familiar one, however, because
of satellite latency on international calls and strained reception
on mobile phones. The delays and dropped calls in the mobile
environment are, in fact, more marked than those stemming
from VoIP use, and they are tolerated by users as the price for
mobility and convenience. Moreover, mobile phones are the
default level of QoS expectation in most developing countries,
where mobile telephony is the dominant voice medium. This
may be particularly true in Africa, for example. And although
VoIP call quality is still inferior to analogue or circuit-switched
systems, many PSTNs in developing countries offer call quality well below what is experienced in developed countries,
anyway. Seen in this light, QoS is a relative concept.
Delay is still a major issue for digital voice transmission,
but other parameters need to be included in QoS for voice
transmission evaluation. The combination of these parameters
will therefore define the end-to-end quality:
–
Jitter, which is the variation in the time between packets
arriving, caused by network congestion, timing drift, or
route changes. A jitter buffer can be used to address this;
–
Packet loss, which introduces audio distortions;
–
Speech coding and decoding, which generate an approximation of the original signal.
6.5.1.2
QoS in Practice
From the consumer’s perspective, QoS is really about a
series of trade-offs involving lower costs and other advantages.
Extending the comparison with mobile phones, consumers
appear willing to pay more for the added value of mobility
than for fixed-line service. Similarly, a consumer using VoIP is
often willing to trade call quality for lower-cost (or free) calls.
In developing countries, grey market users are making exactly
the same trade-off.
From this perspective, QoS for VoIP seems more like a
function of market forces, and in fact, many operators make
great efforts to maintain the QoS at the highest possible level,
so as to attract subscribers. If consumers are unhappy with the
quality of any particular service, they will cease to use it. Many
VoIP applications may not normally meet the mission-critical
voice-quality standards required by corporate clients – unless,
of course, service-level agreements are in place to provide (for
a price) higher-level quality performance and network integrity.
If VoIP QoS appears to be a matter of choice, trade-offs
and market demand, it begs the question of whether QoS is
a major issue. Still, regulators are examining it, particularly in
developing countries, where VoIP is emerging as a low-cost
communication solution.
6.5.2 Regulating Network Integrity and Security
6.5.2.1
Network Integrity
The term network integrity refers to the inherent reliability
of a network and its resilience to external threats, such as natural disasters or malicious acts. The main regulatory issue arises
from the fact that VoIP can be provided independently of the
106
underlying network access. This network/service independence has several implications:
• VoIP can be provided over an access network without the
operator of that network being aware, or having any control over, the voice service provided;
• VoIP can be provided over an IP network using any access
technology;
• VoIP can be provided over an IP network at any location.
This fundamental independence from network operations
has the potential to raise complications for regulators, because
existing network integrity requirements usually were developed when the network and the service were not independent.
VoIP services have introduced the possibility that calls can be
provided independently of the access network provider.
The European Union’s network integrity requirements
are set out in the Universal Service Directive, which calls upon
member countries to take “all necessary steps” to ensure the
integrity of the public telephone network at fixed locations.
The member countries are supposed to ensure that providers
of publicly available fixed telephony “take steps to ensure uninterrupted access to emergency services.” This raises questions
about how VoIP services fit into these requirements:
• Are “nomadic” VoIP services (ones that can be used over
any IP access network) provided at a “fixed” location?
• What are the implications of network/service independence?
• What are “all necessary steps” to ensure the integrity of
the public telephone network at fixed locations in a VoIP
context?
These issues are under discussion in European countries.
Thus far, there is no common European regulatory approach
covering VoIP and network integrity.
6.5.2.2 Electronic Communication Security
Electronic communication security is a broad subject that can
be used to address a large variety of issues. There are eight
dimensions of network security:
• Access control,
• User authentication,
• Non-repudiation,
• Confidentiality,
• Communication security,
• Data integrity,
• Availability, and
• Privacy.
On application layers of an IP network, the operation of
each application (web browsing, e-mail, domain names) brings
with it its own security questions. In each case, specific actions
are taken to minimize risks, such as filtering software for email.
The PSTN is, by nature, a closed network with controlled security and privacy. An IP network is based on open
network architecture, with resulting communication security
vulnerabilities that can affect VoIP as much as any other service.
C HAPTER 6
Box 6.6: SPIT: A Looming Issue
An issue that has recently begun to emerge and may require further consideration is “Spam over Internet Telephony,” or
“SPIT.” Related to the larger problems of Spam and “SPIM” (Spam over Instant Messenger), SPIT is essentially junk mail in
voice form – unsolicited voice messages and unwanted advertising or marketing. This problem has a unique element, in that VoIP
has a broadcast capability. Views differ on the future scope of the problem. Some network security software products are being
developed to incorporate SPIT-blocking technology in future releases. But some operators and analysts are less concerned about
SPIT, because messages have to be streamed to a network, as opposed to simply being mass emailed.
At the same time, others have noted that standard content filters used for spam would be very difficult to apply to voice. The
variability of phrases and pronunciation make algorithms difficult to write, and the technology lends itself to extremely cost-effective solutions for telemarketers. While SPIT is just barely emerging, it is not a new problem. It can be addressed as part of a general
approach to voice security in the IP space that addresses multiple problems and vulnerabilities – including viruses and “denial of
service” (DoS) attacks – that have been identified with Session Internet Protocol (SIP).
Box 6.7: Threats to VoIP Networks and Publicly Available Services
• Distributed Denial of Service (DDoS) attacks threaten the availability of the VoIP network by flooding it with unnecessary data
or attacking the key network elements. DDoS attacks are typically launched from a large number of compromised client
machines and are difficult to defend against in the light of VoIP QoS requirements;
• Thefts of call information can occur by breaching vulnerable VoIP signalling servers. The call information can be as valuable
as the content, so it is likely to be a target for attackers;
• Conversation eavesdropping or recording can occur by breaching VoIP network gateways or other key network points.
Software plug-ins required to sniff out VoIP traffic are available on many open-source websites for free;
• “Call hijacking” or “man-in-the-middle” attacks can occur. These scenarios involve rerouting the connection or modifying
call parameters;
• Identity “spoofing” can occur through manipulating Caller ID;
• Attacks can be made against the terminal equipment software, devices or network servers themselves. The software on these
devices can be vulnerable to the same types of vulnerabilities that affect all operating systems software.
Box 6.8: Defense Mechanisms against Security Attacks
Communication security depends on both the actions of the end user and the security practices of the VoIP service provider.
Security is always a compromise between the utility and cost of the service and the protection mechanisms that are available. In
order to mitigate security risks, the VoIP service provider should consider at least the following measures:
• VoIP networks should be logically separated from other IP services and applications;
• VoIP servers should be hardened and treated using the same security precautions as any other servers that contain confidential information and offer network services;
• VoIP networks should be redundant, in order to ensure the availability of the service. The VoIP network has to be resistant
to denial-of-service attacks. This is especially essential for emergency services;
• Encryption of VoIP traffic can be used whenever reasonable. Encryption can be implemented on the application, transport
or network levels;
• Network devices should be configured properly to restrict unnecessary traffic toward VoIP systems and to ensure the operation of VoIP services.
Box 6.8 discusses some mechanisms that can be used to guard
against attacks.
6.5.3 Lawful Interception
There are times when every nation’s law enforcement
authorities may have justifiable and legal cause to intercept
C HAPTER 6
telecommunication traffic. The regulatory aspects of lawful
interception of VoIP services are complex. In the future it will
be vital for law enforcement agencies to be able to monitor
and intercept internet-based voice traffic, but the VoIP services sector is much more fragmented than the large telephony
operators and involves multiple types of network platforms.
107
The issue of how to manage lawful interception is likely to
become more urgent, as operators move to “next-generation
network” architectures.
Several countries are considering measures to give their
security services powers to intercept e-mails and monitor
internet traffic. Increasing pressure is being placed on national
regulatory bodies to make sure that operators enable surveillance and retain call and traffic information. As more voice
traffic moves to IP-based networks, the same pressures will be
brought to bear on VoIP providers.
For example, in August 2005, the United States FCC
ruled that providers of certain broadband and interconnected
VoIP services must be prepared within 18 months to accommodate law enforcement wiretaps. Under the Communications Assistance for Law Enforcement Act, or CALEA, the
FCC has determined that VoIP services can essentially replace
the telecommunication services currently subject to wiretap
rules, including circuit-switched voice service and dial-up
internet access. As a result, any VoIP provider interconnected
with the PSTN must be wiretap-ready by early 2007.21 CALEA
requires the Commission to preserve the ability of law enforcement agencies to conduct court-ordered wiretaps in the face of
technological change.
Wiretapping, however, raises serious concerns regarding individual privacy rights. In terms of lawful interception,
regulators might play a useful role in helping to determine the
balance between the rights of the individual citizen and the
requirements of government to monitor traffic. At a practical
level, regulators can also help find a balance between the obligations and requirements of law enforcement agencies and the
needs of service providers. Below are some of the issues that
are raised by lawful interception obligations:
• Costs: The cost of complying with wiretap obligations can
be significant. In some countries, the government shares
the costs of lawful interception with smaller operators or
service providers, but where these arrangements are absent,
regulators need to be sensitive to the fact that for smaller
ISPs or VoIP service providers, the cost of purchasing the
108
•
•
necessary equipment to be able to provide access to law
enforcement agencies can be prohibitive.
Area of Responsibility: Another potential problem area is
delineating responsibilities for implementation and compliance between the regulators and law enforcement
agencies. This can lead to difficulties in establishing technical specifications, determining service provider responsibilities and applying remedies for non-compliance.
Standards: Lawful interception, especially of cross-border
services, is highly dependent on standardization bodies
such as the European Telecommunications Standards
Institute. Unfortunately, although standards for lawful
interception in traditional circuit-switched networks are
well defined, there is still a long way to go before interception standards for VoIP are standardized.
6.6 Conclusion
The implementation of regulatory approaches to VoIP will
remain highly uneven for many years to come. In fact, they
may never be uniform. But there is at least a consensus that
regulators should address IP-based networks and services. The
countries that have begun to do so are generating precedents
and providing guidance for others to follow.
It appears that at least minimal additional regulation
of VoIP may be required to ensure quality, security, network
integrity, interconnection, access to emergency services and
further competition in global telecoms markets. VoIP services
offer a truly exciting technological development that may yet
unlock affordable communication solutions for much of the
developing world. Regulators can act to ensure that they assist
in this common goal.
VoIP is a particularly important opportunity for developing countries to provide voice and other services more cheaply
than with traditional PSTN networks. The increased availability of cheaper services will broaden access to a larger number
of citizens, providing another avenue for closing the Digital
Divide.
C HAPTER 6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Clayton Christensen, The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail (New York: Harper Business, 1997). Christensen coined the term
“disruptive technologies,” which are said to satisfy existing customer needs at a drastically lower cost and are simpler and easier to use than previous ones. Disruptive
technologies cannot at the point they are introduced into the market, compete against the traditional products such that they can acquire a large market share.
See http://www.telegeography.com/press/releases/2005-05-31.php
For example, from May 2005, Senegal’s incumbent operator, Sonatel, cut its international call rates to USD 0.19 a minute at weekends and USD 0.31 cents during
office hours. Sonatel was one of early adopters of VoIP for international calling.
In South Africa, as part of its tariff rebalancing, the monopoly operator Telkom reduced international calls overall by approximately 36 per cent over 2004-2005.
Telkom’s licence required it to ensure tariff rebalancing was completed by the end of its five-year monopoly.
See http://www.20min.ch/tools/suchen/story/27383201.
It is also noteworthy that the establishment of the World Trade Organization (WTO) and the GATS Framework Agreement has facilitated a much wider and more
complex exchange process, and facilitated a regime change from one typified by an exchange of traffic, to one characterized by (multilateral) trade in services, and a
consequent shift away from individual, bilateral country negotiations for market access and network interconnection.
The Pakistan carriers use VoIP as licensed long-distance or local loop carriers.
Chinatechnews, “MII: No Plans Soon To Lift Ban Over VoIP”, July 22, 2005, online at http://www.chinatechnews.com/index.php?action=show&type=news&id=2813;
Chinatechnews, “China Netcom Prepares To Crackdown On Illegal Phone Kiosks”, June 20, 2005, online at http://www.chinatechnews.com/index.php?action=s
how&type=news&id=2716
IP-Enabled Services, WC Docket No. 04-36, Notice of Proposed Rulemaking, 19 FCC Rcd 4863 (2004)
“Regulatory framework for voice communication services using Internet Protocol”, Telecom Decision CRTC 2005-28, 12 May 2005.
Local VoIP services are defined as those which use telephone numbers that conform to the North American Numbering Plan and provide universal access to and/or
from the Public Switched Telephone Network (PSTN), along with the ability to make or receive calls that originate and terminate within an exchange or local calling
area.
“Local Competition”, Telecom Decision CRTC 97-8, 1 May 1997.
Memorandum Circular, No. 05-08-2005, Subject: VOICE OVER INTERNET PROTOCOL (VOIP), National Telecommunications Commission of the Philippines (last visited at http://www.ntc.gov.ph/whatsnew-frame.html)
REGULATION OF INTERNET PROTOCOL (IP) TELEPHONY, Statement of the Telecommunications Authority, 20 June 2005, OFTA http://www.ofta.gov.
hk/en/tas/ftn/tas20050620.pdf
China’s government is considering banning the use of unregulated VoIP services. Report in Fierce VoIP, 13 September 2005.
Intrado has introduced a new service that will enable VoIP providers to offer 911 services in compliance with the recent FCC mandate. The company claims that its
V911 Mobility Service will let wireless VoIP providers nationwide position their services as primary line replacements. The solution is designed to accommodate each
jurisdiction’s 911 regulations while supporting static, mobile and out-of-area phone numbers. Intrado supplies 911 integration by transmitting routing instructions
for the local 911 service through the wireless VoIP provider’s modem. The service redirects VoIP 911 calls over the existing wireline service and offers live call-center
support 24 hours a day. See http://lists.fiercemarkets.com/c.html?rtr=on&s=69l,f8zf,lcs,bhks,bxd0,jti9,1419
“Decision on 9-1-1 Emergency Services for VoIP Service Providers”, Telecom Decision CRTC 2005-21.
Federal Communications Commission, Federal-State Joint Board on Universal Service Seeks Comment on Proposals to modify the Commission’s Rules relating
to High Cost Universal Service Support”, FCC 05J-1, CC Docket No. 96-45, 17 August 2005.
Telegeography, “Telof to launch VoIP in unserved areas”, May 16, 2005. Available at http://www.telegeography.com/cu/article.php?article_id=7203
The list is provided in the European Commission’s recommendations.
Federal Communications Commission, First Report and Order and Further Notice of Proposed Rulemaking, FCC 05-153, ET Docket No. 04-295 RM-10865, 23
September 2005. Available at http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-05-153A1.pdf
C HAPTER 6
109
7 STEMMING THE INTERNATIONAL TIDE OF SPAM
Author: John G. Palfrey, Jr., Executive Director, Berkman Center for Internet & Society
and Clinical Professor of Law, Harvard Law School
The anti-spam laws enacted around the world so far have
been largely unsuccessful in stopping spam.1 In almost every
instance, anti-spam statutes have been directed at sanctioning
spammers for their bad acts. An increasing number of countries and other jurisdictions have created such laws or applied
to spam their existing, generally applicable laws concerning
data protection, consumer protection, and protection against
fraud. Yet, in many cases, these laws have missed their target
entirely, with no perceptible impact on actual spammers. Even
worse, the laws have often had negative side effects, in the
form of transaction costs, administrative costs, and a chilling
effect on legitimate senders of e-mail.
No matter what kind of law is enacted or applied, antispam measures require well-conceived, targeted, and coordinated enforcement mechanisms in order to be effective.
Without a doubt, anti-spam investigations are invariably complicated and expensive, presenting challenges for any country seeking to enforce anti-spam laws. Even the U.S. Federal
Trade Commission, with its substantial resources, has brought
only approximately 70 cases against spammers. For developing
countries that have limited human and financial resources for
such work, anti-spam laws can be rendered nearly meaningless
because of the enforcement challenge.
Cross-border cooperation and enforcement is not only
desirable, but also essential to spam fighting. But the variety of
anti-spam laws and underlying legal systems on the books of
various countries makes collaboration extremely difficult. The
challenge of fighting spam through law – to be sure, only one
of the potential modes of regulation – calls for new thinking
and increased emphasis on international harmonization and
collaboration. The only effective means of combating spam
is likely to be a combination of approaches. As noted in the
Chairman’s report of the ITU 2004 Global Symposium for
Regulators (GSR),2 a multi-pronged approach to dealing with
spam is an appropriate measure.
This chapter primarily takes up the question of what
– beyond coordinating with technologists and other countries’
enforcement teams and educating consumers – legislators and
regulators might consider by way of legal mechanisms. First,
the chapter takes up the elements that might be included in
an anti-spam law. Second, it explores one alternative legal
mechanism which might be built into an anti-spam strategy,
C HAPTER 7
the establishment of enforceable codes of conduct for Internet
Service Providers (ISPs). Third, the chapter also examines a
variant of the legal approach where ISPs are formally encouraged by regulators to develop their own code of conduct. ISPs
should be encouraged to establish and enforce narrowly-drawn
codes of conduct that prohibit their users from using that ISP
as a source for spamming and related bad acts, such as spoofing
and phishing, and not to enter into peering arrangements with
ISPs that do not uphold similar codes of conduct. Rather than
continue to rely upon chasing individual spammers, regulators
in the most resource-constrained countries in particular would
be more likely to succeed by working with and through the
ISPs that are closer to the source of the problem, to their customers, and to the technology in question. The regulator’s job
would be to ensure that ISPs within their jurisdiction adopt
adequate codes of conduct as a condition of their operating license and then to enforce adherence to those codes of conduct.
The regulator can also play a role in sharing best practices
among ISPs and making consumers aware of the good works
of the best ISPs. While effectively just shifting the burden of
some of the anti-spam enforcement to ISPs is not without clear
drawbacks, and cannot alone succeed in stemming the tide of
spam, such a policy has a far higher likelihood of success in the
developing countries context than the anti-spam enforcement
tactics employed to date.
7.1 The Spam Problem
The problem of spam is well established. The extent of
the problem is plain to anyone who relies upon electronic mail
(email) for communications. Email and related forms of messaging such as “blogs” (short for “Web logs”) and short messaging service (SMS), have become an important and popular
means of communication in cultures around the world. These
services are cheap, they have global reach, and they are playing
a key role in the development of e-commerce. The proof of
their value is found in their extraordinary global adoption rate,
whether in the form of an e-mail client (such as Microsoft’s
Outlook, Eudora, Thunderbird, or others) or hosted services
(such as Microsoft’s Hotmail, Outblaze, Yahoo! Mail, Google’s
Gmail, Wanadoo or Noos in France, among others).
111
But the openness that has made e-mail and its close cousins such tremendously easy ways to connect is also emerging as
their greatest vulnerability. A combination of economics, technologies, and online behaviour norms has made the incremental cost of sending a spam message nearly zero, while promising
senders a profitable potential return.
At first glance, the economics seem baffling. How can it
possibly be worthwhile to send out grammatically challenged
messages about low-cost pharmaceuticals or pirated software
– offers that the vast majority of recipients ignore and quickly
transfer to their “junk mail” folders? Part of the answer is the
tiny marginal cost of sending spam messages. Because they
cost nearly nothing to send, the response rate does not need to
be very high. And it turns out that enough people do respond
to make the endeavour worthwhile to the spammer. Astonishingly, the Business Software Alliance (BSA) has found that 22
per cent of British consumers they surveyed purchased software through spam.3 Rates for the other five countries BSA
surveyed were similarly high. The bottom line is that spam persists because it is profitable. Unless enough consumers become
educated to avoid or reject spam, the best way to reduce spam
may be to raise the risks and costs to the spammer.
Right now, the costs seem to be landing on consumers.
Every major, credible report on this topic suggests that more
than half of the e-mails sent today are spam, and some suggest
that spam comprises between 70 and 90 per cent of all e-mails
sent.4 The costs of this scourge are borne not by the spammers,
but by those who run networks, employers and the individuals who receive the messages. Spammers – and those who use
spam to perpetrate related frauds – take advantage of the open
design of IP networks to render e-mail costly and nearly unusable for some businesses and consumers.
7.1.1 Legislative Responses
The “extremely rapid growth” of spam5 has led to the
enactment of more than 75 specific laws,6 such as the wellregarded Australian law, the United States’ CAN-SPAM Act
of 2003 and comparable legislation in several dozen countries
around the world.7 These laws have, to date, been unable
to stop spam. Accounts vary somewhat in terms of rates of
growth, but there is no persuasive evidence that the growth of
spam has abated in the wake of anti-spam legislation.8 In fact,
most indicators point in the other direction.9
Spam is best viewed not as a nuisance, but in the context
of cybersecurity. Spam is bad enough as a drain on productivity
and a daily annoyance. But few people consider that spam is
enormously costly to ISPs and others who maintain the network at various levels. Meanwhile, its negative impact is growing by virtue of the bad things it brings with it. Spam is the
preferred delivery mechanism for a range of Internet security
threats: viruses, “phishing” and “pharming,”10 scams with endless permutations, and advance fee frauds, to name a few.11
Spam is also undercutting the efforts of developing countries
to persuade new users to rely on digital communications.
Bill Gates, who is arguably the world’s most powerful
technologist, promised to lead the charge against spam and to
end it within two years of the January 2004 World Economic
112
Forum meeting in Davos, Switzerland.12 He is not alone in
having fallen short in this goal. In fact, most major, well-intentioned ISPs and e-mail service providers, along with many
technology start-ups, have devoted many millions of dollars
to spam-fighting measures. Standards bodies have sought to
improve protocols to snag more spam. User education campaigns have been launched. And governments around the
world have come together to enforce their spam laws and to
cooperate more effectively with one another. The problem
continues despite these many efforts, suggesting that new solutions must emerge and that existing efforts must be better pursued and coordinated.
Some of the most effective recent efforts have been those
lawsuits undertaken by ISPs under a private right of action in
spam legislation. In the United States, the CAN-SPAM Act of
2003 enables ISPs to sue spammers directly. AOL, Microsoft,
and Earthlink – very large-scale providers of electronic messaging services – have each brought actions under this statute,
as well as under state-level computer crime and common law
statutes. This has resulted in multi-million-dollar judgments
and settlements against “spam king-pins” who abuse their networks.13 Microsoft won a USD 7 million judgment that may
well have put an end to one spamming operation that allegedly distributed more than 38 billion unsolicited messages per
year.14
These lawsuits – although few and far between, and
limited to certain jurisdictions – represent a ray of hope that
enforcement by ISPs, with help from customers, might get
the job done against spam. Indeed, the success of these efforts
suggests that ISPs could become the most valuable players in
the effort to end spam. The challenge for lawmakers is how
to create a fair, effective regulatory regime that takes advantage
of ISPs’ ability to help end spam without placing an undue
burden on law-abiding companies.15
7.1.2 A Model Law: One of Several Ways To
End Spam
7.1.2.1 A Combination of Approaches Is Needed
The persistence of spam problem has led policy-makers,
technologists, academics, and many others to come up with
a wide range of possible strategies to end it. The least intrusive approach, most consistent with the end-to-end principle
of network design, is to leave the job to end users, through
simple technologies such as spam filters on e-mail clients. The
improvement of authentication, accreditation, and identity
management technologies ought to help make user-level controls more effective over time.16 At Davos in 2004, Mr. Gates
described Microsoft’s pursuit of solutions to complement
these user controls.17 One approach calls for a combination of
law, code, markets, and norms.18
Meanwhile, the chairman’s report of the ITU Thematic
Workshop on Countering Spam in 2004 contains a range
of proposals, suggesting an intersection of many methods of
spam-fighting.19 This comprehensive, five-part approach calls
for a combination of:
C HAPTER 7
•
•
•
•
•
Strong, enforceable legislation;
The continued development of technical measures;
The establishment of meaningful industry partnerships,
especially among ISPs, mobile carriers and direct marketing associations;
The education of consumers and industry players about
anti-spam measures and Internet security practices; and,
International cooperation among government, industry,
consumer, business and anti-spam groups, for a global
and coordinated approach to the problem.
In fact, virtually every major report on spam calls for a
combination of approaches to combat the problem, rather than
a single, “silver-bullet” solution. This chapter does not take up
in detail each of these anti-spam tools, but rather focuses on
legal strategies, emphasizing those that are relevant to developing countries.
Anti-spam laws are perceived today to be a necessary tool
for all countries, if for no other reason than that they help
facilitate international cooperation in combating spam. Even
the most ardent supporters of user controls and market solutions agree that governments have a role to play in tracking
down and punishing the worst offenders, such as those who
use spam to commit fraud. The existence of interoperable
anti-spam laws creates a common baseline for international
enforcement. A developing country may not be able, by itself,
to enforce its anti-spam law, but that law can provide the basis
for regional and multinational enforcement actions.
A country with experience enforcing anti-spam legislation
may wish to provide human resources to conduct an anti-spam
investigation and enforcement action that leads to another
country. In the absence of anti-spam legislation, however,
such international cooperation is not possible on a systemic
basis. Anti-spam laws are increasingly viewed as one of several
necessary tools for most countries.
7.1.2.2 The Effect on Developing Countries
Spam is arguably a bigger problem in developing countries
than in wealthier countries, where anti-spam mechanisms are
more robust. Many developing countries do not yet have antispam laws,20 and those that do often do not have resources to
enforce them.21 Meanwhile, the effects of spam are often relatively more costly in developing countries. ISPs are frequently
deluged by spikes in spam, which lead to network slowdowns
and breakdowns.22
Moreover, many people in developing countries send
emails from shared Internet connections and equipment, such
as at cybercafés or other public access centres. These services
ordinarily rely on hosted email services with limits on inbox
sizes. Accessing email becomes too expensive if per-minute
charges paid to cybercafé owners are consumed by cleaning
spam from their inboxes. Even worse, legitimate emails are
bounced because the limited space of their inboxes is consumed by spam.
Officials from developing countries often point to the fact
that most spam still comes from the United States and other
wealthy countries, which have done little to help developing
C HAPTER 7
countries cope with the problem. In addition, they note that
the resources of regional bodies such as the OECD are not
consistently available to developing countries. This leaves them
at a comparative disadvantage in fighting spam.
The answer for developing countries is not simply to copy
anti-spam laws enacted in developed countries. That approach
is unlikely to be effective. Anti-spam laws aimed at sanctioning spammers may be of little use in developing countries if
the spammers are outside their jurisdiction. The challenge is
to tailor legislation to patterns of usage in developing countries
and to consider all avenues to combat spam, such as implementing enforceable codes of conduct for ISPs.
7.1.3 An Alternative Mechanism: Enforceable
Codes of Conduct
In addition to enacting anti-spam legislation, developing
countries could require ISPs to establish an industry code of
conduct on spam. The enabling legislation for such a code
could stipulate that the nation’s regulatory agency would
enforce the code against any ISP that materially violated it.23
Such a proposal cuts jarringly across the grain of most internet regulation to date. As essential players in developing ICTpowered economies, ISPs have generally been left alone by
legislatures, administrative agencies, and judges. They may be
licensed and overseen by regulators in some contexts, but ISPs
have largely been immune from prosecution for bad acts committed by people through their services.
7.1.3.1 Elevating the Role of the ISP
Ideally, it is not an ISP’s job to be a gatekeeper. The ISP
should pass all packets from sender to receiver, with end users
deciding what to send and what to receive. Any departure from
this model should be undertaken only when serious circumstances warrant it. In addition, regulation should be handled
with a light touch, and any burdens placed on ISPs should not
be starting points for more intrusive regulation.
It is essential to acknowledge how the internet has changed
since its inception. We use the network far differently than
any of its early architects could possibly have imagined. The
“community” of users is now more far-flung than it ever was,
and they no longer expect to know one another, as the earliest
academics and military users did. The internet’s architecture
is a victim of its own success. The conventional wisdom that
no intelligence should be built into the heart of the network
– the so-called end-to-end principle – is still held dear by many
technologists, but it is no longer fully reflected in reality. A
large number of control points have been built into the network – often to deal with massive problems like spam.24
ISPs still enjoy broad immunities in many jurisdictions
from claims based on what others do on their networks. For
example, they rarely face copyright violation or defamation
claims. But they are increasingly called upon to play a role in
protecting and policing the internet. There are substantial risks
associated with placing such jobs in the hands of ISPs – particularly to civil liberties – so any legislation that mandates a
greater supervisory role must be carefully drafted so as to mitigate these risks.
113
7.1.3.2 Establishing an Industry-Led Approach
Countries should work to establish an industry-led regulatory approach that provides a mechanism for regulators to
step in against the worst spam abusers. This proposal is not
meant to presage a wholesale shift in the role of ISPs. Nor is
it meant to indicate a rejection of the end-to-end principle as
a preferred design matter. ISPs already bear the brunt of the
costs of spam. The role of the law and the regulator should not
be to burden ISPs further, especially given the constraints they
already face.25
Rather, the goal is to reduce spam in a way that protects
responsible ISPs. As the internet has developed into a complex
network of networks, ISPs are positioned, for good or ill, as
key gatekeepers. ISPs that implement responsible, effective
anti-spam measures, while preserving the civil liberties of their
users in a manner that is consistent with local law, should be
rewarded for their good behaviour. One means of rewarding
those responsible ISPs is for regulators to hold their irresponsible competitors accountable. This would create a level playing
field for responsible ISPs.
ISPs are no strangers to fighting spam. ISPs around the
world have taken an active role in attacking spam at the source,
before it clogs their customers’ inboxes. Anti-spam measures
implemented by ISPs cover a wide range. Many ISPs participate in industry-wide working groups, such as the Messaging
Anti-Abuse Working Group.26 Many also work with standardsetting organizations developing technical solutions.27
ISPs’ initiatives are often geared toward improving security
and decreasing the vulnerability of users and of their networks.
When they succeed, it can often be a strong selling point for
them. For example, Google’s Gmail, a free Web-based e-mail
service, removes hyperlinks from messages that the service
believes to be phishing attempts.28 The large U.S.-based ISP
Earthlink requires all e-mail messages to be routed through its
mail servers, in order to reduce the impact of “zombie” networks. Earthlink also mandates that users’ e-mail programs
submit passwords to transmit messages.29
While these methods can reduce the burden of spam,
their effect is minimal if consumers do not also take steps at
the “client” level of the network. End users may not update
their own virus software automatically or regularly. Or, they
may download programs that contain “malware” and “spyware”
that compromise their computers, posing a risk not only to
themselves but to other users worldwide, since their PCs may
be hijacked to relay spam to other unsuspecting consumers.
Governments and ISPs both have incentives to end spam.30
ISPs bear a large amount of the cost of spam and get nothing
in return – unless they are charging a premium to spammers
in exchange for sending spam out on their behalf. ISPs also are
relatively close to the problem. After all, spammers need ISPs
to get access to the internet to dump their messages. While
spammers are increasingly sophisticated in evading tracking,
a concerted effort among cooperating ISPs (and possibly law
enforcement officials and end users) can find the worst offenders. The routing of spam can be traced and mapped at a network level.31 While ISPs are often short on cash flow, many do
114
have the financial and human resources to play a key role in
the anti-spam fight.
National laws can mandate the development of codes of
conduct by and for ISPs. Adherence to the code could be a
licence condition, or it could be implemented through a rulemaking proceeding, via a common set of regulations that applies
to ISPs whether licensed or authorized, much as operators are
required to provide interconnection, the rules for which are
spelled out in interconnection regulations with industry participation. The law would give ISPs the first opportunity to craft
the code, outlining acceptable behaviour for ISPs and their
customers. Preferably, the code would prohibit spam, phishing,
spoofing on the ISPs network, and similar practices. It could
also suggest or endorse the best use of spam filters and other
technological tools for customers and ISPs to fight spam. The
regulatory agency would approve and, in many cases, enforce
the code.
Under such codes, ISPs would commit themselves to
denying service of any kind to spammers, phishers, spoofers
and other bad actors who violate these policies. Such codes of
conduct would be led by industry and made functionally consistent among all players across the industry, but as part of a
process that is grounded in law and provides a role for regulators. The regulator would be empowered to approve the code
and to enforce the code if the ISP deviates from its terms in
material fashion.
Regulators are better able to do their job under this scenario, as compared to the straight enforcement role against
spammers, since the regulators would primarily interact with
ISPs. The ISPs are largely running legitimate businesses, are
incentivized to help solve the problem (so long as they are not
cheating), and are easy to find relative to the spammers, who
are often not in the same country and are constantly hiding
behind technological smoke and mirrors. The ISPs, in turn,
would be responsible to keep tabs on those customers who are
engaged in illegal activity and to spurn offers for premium payments to provide spammers with an onramp to the internet.
This mechanism would empower the regulator to apply a
default code of conduct where ISPs fail to develop one or until
an acceptable policy is set forth by the ISP. Such a mechanism
would also include the regulator’s certification of the code
which ISPs could use in their advertisements, to ensure customers that the ISP is taking all available steps to protect its
customers, and the network at large, from spam. The system
would also involve a reporting mechanism so that victims of
spam, phishing, spoofing and the like can report such activity
either to the ISP or the regulator for follow-up investigation
and action.
An enforceable code of conduct is not without drawbacks.
The code must be narrowly tailored to curb spam and related
bad acts. It should not be used as a back-door measure to overburden ISPs, such as by:
• Imposing anti-spam obligations where no technical solution yet exists (as with many anti-spoofing requirements);
• By using anti-spam measures as a means to limit legitimate political discourse or other protected speech; or
• By infringing on the privacy interests of citizens.
C HAPTER 7
It is essential that the industry develops and approves the
code of conduct – or, at a minimum, collaborates with regulators in this task. Industry “buy in” is important, because the
code will require frequent updating to reflect new developments in spamming practices and anti-spam technologies.
7.1.3.3 Voluntary Codes of Conduct
As an alternative to a mandated code, enforced by regulators, governments might encourage ISPs to develop their own,
industry-enforced codes of conduct. In fact, many ISPs are
taking this step without any encouragement. Terms and norms
are often built into “acceptable use” policies for customers and
peering arrangements.32 Under this voluntary model, regulators could advise the industry in developing the codes. It could
then help consumers find the ISPs that have developed or
signed on to those codes. If a vibrant ISP market emerges, consumers could then choose ISPs that have proactively tried to
fight and reduce spam.
Finally, regardless of whether ISPs are compelled to establish codes of conduct or do so voluntarily, regulators have
an important role to play in educating and raising awareness.
Consumers, businesses, ISPs and cybercafé operators need
information on technical solutions such as spam filters, as well
as warnings about viruses and fraudulent activities that have
been detected. There is much to be gained from governmentindustry collaboration in protecting consumers from spam.
7.2 An Outline of a Model Law
7.2.1 The Context for a Model Anti-Spam Law
Representatives of many countries, particularly in developing regions of the world, have sought a model law for
combating spam. The topic was discussed intensively at two
international gatherings hosted by ITU. The first, held in the
summer of 2004, was devoted to the issue of spam, while the
other, a year later, focused on cybersecurity. This chapter draws
upon the many resources developed to date, in an attempt to
create a model anti-spam law. There are multiple potential
benefits of such a document:
• Clear guidelines – Email senders that want to comply
with legal requirements could more easily learn what rules
apply to them and could then follow them more consistently.
• Jurisdictional Consistency – Enacting a similar, model
law in many jurisdictions would free ISPs and email senders from having to attempt the near-impossible task of tailoring messages for recipients in different jurisdictions.
• Easy adoption – Legal systems that do not yet have
laws governing spam would have a ready-made model to
implement, reducing the burdens of drafting, implementation, and coordination.
• Enhanced enforcement – Regulators could enforce laws
more effectively and easily since their systems would share
harmonized definitions of offences, burdens of proof, and
C HAPTER 7
•
•
•
exceptions. Greater harmonization would make broadbased cooperative arrangements more likely to arise.
Stronger norms – Broad international consensus on the
meaning of spam, and what constitutes unlawful abuse of
electronic communication, would strengthen norms that
deplore such conduct.
Fewer havens for spammers – As more governments
adopted the model law, spammers would have fewer
friendly locations to establish operations. This would
increase their costs and reduce the financial incentives to
engage in massive spamming.
Increased sharing of best practices – Since legal systems would share harmonized provisions, regulators and
enforcers could more easily collaborate to develop and
share best practices for implementing spam laws.33
Even well-crafted anti-spam laws, implemented in every
jurisdiction, will never get the job done alone. But anti-spam
legislation can be a useful element of a coordinated anti-spam
strategy. A good anti-spam law should distinguish between
good actors and bad actors and mete out punishment accordingly. Moreover, if spammers were liable for each spam message they send, the level of fines would increase exponentially,
according to the scale of the spam operation.34 Enforcement is
the key – and the most difficult element – particularly in developing countries.35
The development of a model anti-spam law should be
collaborative and inclusive. As with any model law (or any official document with the force of law) an anti-spam law must be
flexible enough to dovetail with existing laws, including antifraud, consumer-protection, telecommunication and internet-specific laws and regulations. One relevant example is the
process that the United Nations Commission on International
Trade Law (UNCITRAL) undertook in establishing its Model
Law on Electronic Commerce (1996).36 UNCITRAL’s e-commerce model law does not specifically address spam, which did
not exist as in 1996 as the huge issue that it is today. Anyone
designing an anti-spam model law should also consider the
broad range of laws on the books today in many countries,
containing variations that are worth considering but that are
too numerous to be included in this chapter.37
Most of the existing anti-spam laws are directed at controlling spammers’ behaviour. This seems appropriate, since
spammers directly cause the problem. But the current slate of
laws has failed even to curb the growth of spam, much less to
reduce the problem.38 Why have they failed? Some observers
argue that the countries generating the largest proportion of
the world’s spam have done too little at home to stop the problem.39 Those making this argument especially criticize reliance
upon “opt-out” rules that allow spam unless consumers specifically ask not to receive it. Even then, opt-out rules are not
enforced aggressively enough.
It is not enough to blame the greatest spam-producing
nations, though. No country in the world – including those
lauded as the most effective in combating spam – has made
significant inroads using classic enforcement mechanisms. Of
course, it would help if governments updated their laws in
115
light of their apparent inadequacy, but that takes time. Other
observers suggest that anti-spam laws should be focused not
on the spammers themselves, but rather on the (often dodgy)
companies for whom the spam is sent. 40
In addition, since anti-spam statutes can affect civil liberties such as
free speech and personal privacy, definitions may play a pivotal role
in determining whether the statute is permissible under a country’s
constitutional framework or sufficiently protective of citizens’ rights.
The primary issue is that little emphasis is placed on investigation, enforcement powers, or resources. It is not that hard
to build and win a case. Most spammers and their clients eventually can be found, with enough hard work and cooperation.
The problem is that each investigation is so time-intensive and
costly that police and prosecutors often decide that the costs
outweigh the benefits. One of the core tenets of the model law
described below is that it emphasizes creating a framework for
national enforcement, international coordination, and distributed monitoring through the ISP code of conduct.41
The following are some of the key terms to be included in
the definitions section of the model law, (although this is not a
complete list):
•
Annotation: An important question for any anti-spam law is
whether or not to include a prohibition on the use of, or trafficking
in, technologies that support spamming, such as address-harvesting
software. If such as ban is included in the law, the term must be
carefully defined so as to avoid banning useful technologies of general applicability that may be used for address-harvesting. Another
approach is not to ban any technology, but rather to bar its use for
gathering e-mail addresses for spamming.
7.2.2 Elements of a Model Spam Law
The draft model law presented in this section as an annotated outline roughly follows the structure of the Australian
anti-spam law, which is widely regarded as one of the most
well-conceived statutes of its kind in the world.42 This section
describes the key elements of a model law, offering suggestions
for options at each stage of the drafting process.
One threshold issue is whether the law will be an “optin” or an “opt-out” statute. An opt-in statute makes it illegal to
send spam unless a recipient has affirmatively agreed to receive
it. Often, only tacit acceptance is required, such as the existence
of an ongoing business relationship of some kind. An opt-out
statute, on the other hand, permits spam unless the recipient
has specifically informed the spammer that he or she does not
want to receive it.
The decision to choose an opt-in or opt-out approach will
reverberate throughout the law from that point onward. For
instance, in an opt-out system, the provision to establish an
“unsubscribe” function will be more essential and take on a different character than in an opt-in law, which presumes that the
receiver already gave a green light before receiving any spam
messages.
One deficiency of many spam laws is a lack of clear definitions. The draft model law, below, seeks to head off variations
among definitions adopted in different jurisdictions, because
these variations could undermine international cooperation on
enforcement.
•
•
Authorization. The law should clarify what it means for
an individual to authorize sending a message that could be
defined as spam.
Annotation: This definition may take on greater or lesser significance depending on whether the law is designed as opt-in rather
than opt-out. Depending upon the nature of the law adopted and
the use and definition of the term “consent,” this definition might
not be necessary.
•
Commercial. The law must specify with precision what
constitutes a message sent for commercial purposes.
Commercial messages sent to recipients with whom they
do not have a previous commercial relationship are likely
to serve as the core, prohibited type of content..
Annotation: One key issue facing development of a useful
model law is variation in the treatment of speech rights in different countries. In Australia and the United States, for instance,
legislators and regulators have stayed clear of regulating unsolicited
political messages in light of constitutional protections for political
speech. Most anti-spam laws focus not on the content of the message, but rather on the intent of the sender. Spam legislation varies
as to whether or not it applies only to commercial messages, but it is
important to define what constitutes “commercial” in any event.
Section 1: Introduction and Definitions
The law should clarify that it establishes a scheme for
regulating commercial e-mail and other types of commercial
electronic messages.
116
Authority, or Regulator. The law should specify the
entity or individual that has jurisdiction over the antispam law. Countries vary as to the precise placement of
this authority, which might be vested in the telecommunication regulator, the consumer protection authority, the
trade regulator, or another authority.
Annotation: If multiple regulators are tasked with enforcing anti-spam rules, a precise division of responsibilities should be
established, either in the definitions section or, more likely, in the
enforcement-related provisions.
Draft Model Law
Annotation: The introduction section of the law ought to set
forth the definitions, which take on special significance in the antispam context. On the one hand, the terms must be broad enough
to encompass emerging types of spam as they develop. On the other
hand, the provisions must be precise enough to be clearly understood.
Address-harvesting software. The law should define
what types of computer applications used to harvest email addresses are banned under the statute.
•
Consent (or, Affirmative Consent). The law should
clearly state what the recipient must do to signal willingness to receive e-mail from a particular sender. The law
could use the term affirmative consent, which means that (A)
C HAPTER 7
the recipient expressly consented to receive the message,
either in response to a clear and conspicuous request for
such consent or at the recipient’s own initiative; and (B) if
the message is from a party other than the party to which
the recipient communicated such consent, the recipient
was given clear and conspicuous notice at the time the
consent was communicated that the recipient’s electronic
mail address could be transferred to another party for the
purpose of initiating commercial electronic mail messages.
Annotation: This definition should be coordinated with the
definition of the term “authorization,” as needed.
•
Electronic message. The law should specify what constitutes an electronic message. In the Australian statute, an
electronic message is a message sent using (a) an Internet carriage service or (b) any other listed carriage service. Also, an email message is sent to an electronic address
in connection with (1) an e-mail account; (2) an instant
messaging account; (3) a telephone account; or (4) a similar account.
Annotation: An important area to consider is what applications the anti-spam statute covers. The best anti-spam laws will be
general enough to cover ICT-based unsolicited messaging in formats
that have yet to be devised, as well as those that exist today. Short
Messaging Service (SMS) text messages on cellular phones, spam
over the instant messaging protocol (“spim”), web blogs (especially
in the comments fields), spam over Internet telephony (SPIT),
voice messaging over Internet telephony and Really Simple Syndication (RSS) are important current variants of traditional e-mail
spam that drafters may wish to keep in mind.
•
Evidential (or evidentiary) burden (or, burden of
proof). The law should define carefully which party bears
the burden of producing evidence.
Annotation: One of the key problems that enforcement
authorities face is a high burden of proof placed upon the prosecution in instances where they must show conclusively that a user did
not opt-in to receiving spam. Virtually no individual can prove the
negative – that they never entered into a commercial relationship,
or never once hit “OK” in a click-through contract. To place the
burden on the regulator to prove this negative is to hamstring her or
him in the enforcement process.
•
Internet service provider (or Internet carriage
service; Internet content provider; E-mail service
provider; Telecommunications service; or the like
depending upon jurisdiction). The law should define
what type of service the statute covers. The essential part
of the definition is that the covered party provides a connection between an end-user and the internet, for a fee.
Annotation: In many jurisdictions, a wide range of definitions
for ISPs are established by various internet-related laws, so special
care should be taken to harmonize definitions across statutes, for
clarity’s sake. U.S. law, for instance, has more than 40 potential
definitions for terms that resemble “Internet service provider.” 43 The
elimination of ambiguity is particularly important for this model
law, which contemplates setting an affirmative requirement for ISPs
to develop an enforceable code of conduct.
C HAPTER 7
•
Send. The law should clarify that the definition of “send”
includes attempts to send.
Section 2: It is unlawful to send unsolicited commercial electronic messages
Annotation: The scope of what type of message is unlawful to
send, combined with the definition of the terms of what is banned,
is a crucial element of any spam law. Countries vary widely in
terms of whether messages beyond “unsolicited commercial e-mail”
are included under the law. For instance, non-commercial bulk
e-mail is included in the definition of “spam” in some anti-spam
legislation and not in others. This is also the juncture at which
each country must decide whether to join the opt-in or opt-out
camp. Virtually all anti-spam laws focus upon the act of sending
(or attempting to send) as the core, operative offence. An additional
prohibition for this section might be to hone in on the act of paying
someone to send unsolicited commercial electronic messages on one’s
behalf. Some states also bar the sending of unsolicited charitable and
issue-oriented (political) messages, but that step is dangerous and
not advocated here, given the importance of political speech to wellfunctioning government systems.
Section 3: Commercial electronic messages must
include accurate sender information
Commercial electronic messages must include information about the individual (or organization) who (or that)
authorized sending the message.
Annotation: The law might also require that commercial
email be identified as an advertisement, by requiring that “ADV”
or the like be included in the header. The law could also require
commercial email to include the sender’s valid postal address. Some
activists have also called for the requirement that senders label sexually explicit messages in the subject line. The labeling requirement
is hotly contested by e-mail marketers, who fear that ISPs or individuals will filter out all such messages, even if they are legitimate
commercial offers.
Section 4: It is unlawful to include false information
in any commercial electronic messages
Commercial electronic messages must not include false
information. That includes an email’s “from,” “to,” and routing information, which should include the originating domain
name and email address. The subject line cannot mislead the
recipient about the contents or subject matter of the message.
Annotation: Most experts contend that an anti-spam law
ought to contain such a ban on inclusion of false information as
a supplement to other provisions, such as the outright bar against
sending an unsolicited message. Without the general ban on unsolicited emails, this accuracy requirement can be criticized as effectively
permitting spam that is unwanted but accurate. Much of the criticism leveled against the U.S. CAN-SPAM Act of 2003 has followed this argument.
117
Section 5: It is unlawful to send a commercial electronic message without a simple means for recipients
to indicate that the recipients do not wish to receive
any further commercial electronic messages from the
sender
Commercial electronic messages must contain a functional “unsubscribe” or opt-out facility. If a recipient exercises
the right to request no further emails, the sender must be
bound to honour that request. In an opt-in regime, an unsubscribe provision would basically ensure that any recipient who
had previously opted in could reverse that decision and opt out
at any time.
Annotation: In the United States, a sender must provide a
return email address or another internet-based response mechanism
that allows a recipient to ask the sender not to send future email
messages to that email address. The sender must honour that request.
Any opt-out mechanism a sender includes must be able to process
opt-out requests for at least 30 days after commercial email is sent.
When a sender receives an opt-out request, the law allows 10 business days to stop sending email to the requestor’s email address. A
sender may not help another entity send email to that address, or
have another entity send email on its behalf.
Also, it is illegal for a sender to sell or transfer the email
addresses of people who choose not to receive that sender’s email,
even in the form of a mailing list, unless a sender transfers the
addresses so another entity can comply with the law. These provisions, while sensible, are believed to have a very low rate of compliance. Most critics also believe that unsubscribe responses by recipients
are frequently used to bolster spamming lists, since the spammers
then know that the email has reached a real recipient.
Section 6: The use of, and trafficking in, address-harvesting software and the resulting lists of electronic
mail addresses are prohibited.
Address-harvesting software must not be supplied,
acquired, trafficked in, or used. An electronic address list produced using address-harvesting software must not be supplied,
acquired, trafficked in, or used.
Annotation: There is a wise presumption generally against
banning general-purpose technologies. Any provision of this sort
ought to exempt the makers of general-purpose technologies (for
instance, a spreadsheet or software enabling a user to write a simple
program that could scrape information from the Web) that might
be used by spammers to harvest e-mail addresses. The law might
also include a prohibition against hacking into databases of e-mail
addresses, although in many jurisdictions such acts would be covered under statutes related to computer crimes, larceny, trespassing
or other offences.
Section 7: Remedies include civil penalties, injunctions,
and criminal penalties
The main remedies for violation of the law would be civil
penalties and injunctions. Criminal penalties, including imprisonment, are also sometimes sought when false representation,
118
use of another’s computer to perpetrate a fraud, or similar acts
are involved.44
Annotation: The law might also include a provision making
it a criminal offence for an ISP knowingly to accept premium payments from spammers who use the ISP’s network to send their
spam. Similarly, the law might include a provision that makes the
knowing hiring of a spammer to send out unsolicited commercial
e-mail a criminal offence.
Section 8: Causes of Action
This section would establish a cause of action for regulators against anyone hiring a spammer to distribute bulk email
for them (i.e., the owner of a website to whom a spammer is
paid to direct traffic, or the party seeking to drive up the value
of a certain equity offering, etc.)45. The law might also include
additional causes of action, enabling ISPs, enforcement officers
in lower jurisdictions, and harmed individuals to initiate cases.
Section 9: International Cooperation
The law should create a mechanism for international information sharing and, possibly, formal cross-border enforcement
support. These rules would simplify the process for exchanging information and encourage exploration of memoranda of
understanding (MOUs) and similar means of cross-border
cooperation.
Annotation: Much of the emphasis of far-sighted regulators in
recent years has been on improving cross-border enforcement efforts.
The U.S. Federal Trade Commission has been encouraging the
U.S. Congress to pass legislation to make such cooperation more
likely to succeed. Consider also the work of the International Consumer Protection and Enforcement Network, which involves dozens
of countries in “sweep days” to rid the internet of scams.46
Section 10: Jurisdiction
An effective anti-spam law might include provisions
designed to assist enforcers by resolving jurisdictional ambiguities.
Annotation: Such a provision could simply clarify what it
means for a message to originate or be received within that country
and how the regulator will treat such situations. On a more elaborate
level, in the United States, the state of Washington’s anti-spam law
established a database that includes many of the e-mail addresses in
that jurisdiction. The purpose is to protect the state’s residents.47 A
list of that nature, held in one place, however, could be an attractive
target for hackers. This concern is mitigated by the fact that spammers apparently do not have much of a problem coming across large
swaths of e-mail addresses through other means.
Section 11: Enforceable Codes of Conduct by ISPs.
An effective anti-spam law might include sections related
to the development and enforcement by regulatory authorities
C HAPTER 7
of industry-derived and implemented Codes of Conduct for
ISPs.48 Such provisions might include:
a) An introduction, explaining the intention to establish such
codes of conduct.
b) A provision granting regulators authority to require all
ISPs to develop a code of conduct for that jurisdiction.
c) A description of the multi-stakeholder process involved in
developing codes of conduct, including what groups will
represent the interests of consumers and industry.
d) A provision establishing a registration process for codes of
conduct.
e) A provision enabling consumers to access registered codes
of conduct.
f) A provision enabling the regulator to draft a code of conduct in the event that industry cannot agree or otherwise
fails to develop one.
g) A provision enabling the regulator to reject a proposed
code of conduct in the event that it lacks appropriate community safeguards.
h) A description of the process for the regulator to issue a
warning to an ISP for apparent breach of the code prior to
taking an enforcement action.
i) A provision granting power to the regulator to enforce the
code in the event of breach by the ISP.
Annotation: A similar structure is set forth in Part 6 of Australia’s Telecommunications Act of 1997 covering industry codes
of conduct (see Box 7.1). There are several issues to be considered,
many of which are set forth in the section that follows. The law
would need to establish a deadline for compliance and provide for
periodic updating of the code. One option would be to task an
industry association (if one exists in that jurisdiction) to develop the
code. The next decision would be whether all ISPs have to comply
with a code developed by the association. The enabling provisions
for the code might allow ISPs to opt out of a code developed by the
association and register a separate code with the regulator, provided
the ISP’s self-developed code sufficiently protects the public interest.
7.3 Codes of Conduct
The primary goal of a code of conduct is to ensure that
ISPs that provide a route to the internet – the source ISPs – are
taking adequate steps to keep spammers off the network. The
effect of the code should be to level the playing field for ISPs
that are actively seeking to rid the network of spam instead of
profiting from sourcing it. While there are many risks in regulating ISPs more extensively than they have been in the past, a
carefully balanced set of provisions will benefit not just customers, but all well-intentioned ISPs, too.49
In virtually all instances, industry knows better than most
regulators what technical solutions to spam exist and can be
implemented.50 Regulators have a role to play in ensuring that
industry does all that it can to put technical and policy solutions in place and to share best practices.
The use of industry codes of conduct is a promising
mechanism that has been under-utilized in the anti-spam fight.
C HAPTER 7
A similar strategy has been used for a variety of other issues,
such as interconnection, number portability, and other technical coordination issues. If combating spam is not in the remit
of the telecommunication regulator, a similar mechanism
could be established for consumer protection authorities, data
protection authorities or other similar bodies. For the purposes
of this chapter, the code of conduct has been included in a
model anti-spam law, but such a set of provisions could easily
fit within other sections of a country’s legal codes, such as the
telecommunication laws and regulations. The code of conduct
does rely, however, upon core elements of an anti-spam statute.
7.3.1 Procedural Steps Toward an Enforceable
Code of Conduct
Industry codes of conduct should be developed in a spirit
of minimal regulation of the internet and as a measure of
private and public sector cooperation to address the growing
problem of spam. The process of drafting a code likely would
include several key steps:
•
The relevant industry member or members are granted
the first chance to develop their own code of conduct,
based upon the stated goals of the enabling law or regulations. The process by which a code is drafted should be
set forth in the law or regulations so as to ensure broad
and open participation by key stakeholders.
•
Where appropriate, the regulator can help by sharing best
practices. This can be done, for example, through the use
of ITU’s Global Regulators Exchange (G-REX)51 or face
to face meetings such as ITU’s annual Global Symposium
for Regulators (GSR). Regulators may also be able to tap
into international resources such as the OECD’s Spam
Toolkit, which is under development. A draft is accessible
at http://www.oecd-antispam.org
•
The relevant industry members present the draft code to
the regulator for its approval.
•
A new body, or an existing regulator with relevant expertise, takes responsibility for the administration and registration of the code.
•
If the industry fails to develop a code, or if the code is not
deemed acceptable, the regulator has the power to step
in to draft or revise it, ensuring that sufficient anti-spam
measures are being taken by ISPs, network operators and
other potential spam carriers.
•
The industry members are expected to enforce the code
against their customers and those with whom they peer.
The enforcement is meant to prohibit the worst acts of
spamming, not to encourage an ISP to monitor messages
any more than they already do. The expectation is that
ISPs would only need to take reasonable measures, such
as investigating when they receive an unusually large
119
•
•
numbers of complaints against a single customer or when
the regulator passes along such complaints.
The regulator or administrator provides a mechanism for
handling end users’ complaints against ISPs for failure to
live up the code.
If industry members fail to enforce the code, the regulator
is empowered to take action against non-compliant ISPs.
Possible sanctions include fines, harsher licensing requirements, or lawsuits.
Annotation: One issue to consider is which parties would
have a right of action to sue a non-responsive ISP. For instance,
consumers who have experienced damage by spam or phishing could
be given the right to go to court to sue ISPs directly for violating
the code of conduct. Also, regulators could require ISPs to include
in their customer contracts binding agreements to honour the code.
This would allow consumers and companies to sue not only under
an anti-spam law, but also pursuant to laws governing breach of
contract.
•
•
The code could also create a “certification” or “accreditation” system, allowing ISPs to publicly advertise their
compliance with the code. The accredited ISPs would be
able to display a “trust mark” signifying their status, helping consumers to make reasonable decisions about which
service to choose.52
The code should also include a mandatory review or
“sunset” provision to ensure that the rules remain effective
and appropriate in a fast-changing technological and legal
environment.
7.3.2 Elements of a Model Industry Code
of Conduct
Like a model law, an industry code of conduct should be
developed in an inclusive, collaborative atmosphere, designed
to elicit the best thinking from a range of experts and concerned
stakeholders.53 The code should set forth the responsibilities of
ISPs and other actors with sensitivity to local concerns. But it
should also take into account the cross-border nature of the
problem. Key elements of a model industry code of conduct
might include:
• A series of common definitions that correspond to the
definitions in the enabling law.
• Procedures ISPs should follow in dealing with obvious
spam that comes into the ISP’s sub-network (including
procedures relating to the provision or use of filtering
software).
• A commitment not to serve individuals or companies that
send unsolicited commercial email in bulk, and to terminate those clients when complaints and subsequent investigations reveal that they have been spamming through the
ISP’s network. This should also include a commitment to
refuse payment, or any enticement of a premium payment,
offered by a known spammer for any service.
• A commitment to give ISP subscribers information about
the availability and use of software for filtering spam at
the client level. ISPs should also commit to helping subscribers prevent their computers from being infected by
120
•
•
worms, “Trojans” and other malware that turns computers into spam “zombies.”
A commitment to assist in developing and evaluating filtering software that gives end users a maximum level of
control over what to accept and to reject.
Suggested best practices that ISPs can implement, as
appropriate, in order to minimize or prevent spam. At
present, such suggested best practices might include some
of those set forth in the London Action Plan.54
The London Action Plan stemmed from a July 2004
meeting of “government and public agencies from 27 countries
responsible for enforcing laws concerning spam.” They generated several recommendations affecting:
• The optimal configuration of servers and other network
devices to minimize or prevent spam;
• A commitment to taking meaningful zombie-prevention
measures;55 and,
• A statement of principles for entering into peering
arrangements only with ISPs that adhere to the full code
of conduct.
The provisions of codes will no doubt change rapidly as
the nature of the problem changes. Today, up to half of all
spam is sent through “zombie” computers, suggesting that it is
vital to help end users prevent the hijacking of their computers.
Once this loophole is closed, spammers are sure to look for
other mechanisms, and codes will have to be updated accordingly. The enabling law should be flexible enough to accommodate changes in the technological landscape.
7.3.3 Hazards of Enforceable Codes of Conduct
Adopting a regime of enforceable codes of conduct for
ISPs is not without hazards. A well-designed policy, however,
should be able to mitigate these risks, which are worth exploring here.
The purpose of industry codes of conduct should be to
give ISPs incentives to exclude spammers from their networks,
not to over-regulate ISPs. Nor should regulators use codes to
deputize ISPs to overzealously block email or monitor conversations. Codes should be strictly limited to requiring ISPs to
shut down spammers. They should not be employed for other
objectives, such as shutting down email with what the government considers unpalatable political messages or for surveillance of a country’s citizens. The risk is that empowering ISPs
as gatekeepers will lead them to avidly look into the nature of
messages sent across their networks.
This potential pitfall points back to the importance of
defining spam in the anti-spam law. A properly crafted law
should rule out abuses of authority in the name of preventing
spam. Regulators should clearly focus on the goal of weeding
out the worst, most obvious cases of spamming, rather than on
pressuring ISPs to shut down legitimate e-mailers.
Another risk in establishing an enforceable code of conduct stems from political realities. In many countries, ISPs have
enjoyed broad immunity from regulation and may oppose any
spam-related responsibilities. More often, the ISP may be a
monopoly, state-owned provider that generates important rev-
C HAPTER 7
Box 7.1: Australia Telecommunications Act 1997 – SECT 117
Registration of industry codes
1) This section applies if:
a) the ACMA is satisfied that a body or association represents a particular section of the telecommunications industry
or the e-marketing industry; and
b) that body or association develops an industry code that applies to participants in that section of the industry and deals
with one or more matters relating to the telecommunications activities or e-marketing activities, as the case may be, of
those participants; and
c) the body or association gives a copy of the code to the ACMA; and
d) the ACMA is satisfied that:
i) in a case where the code deals with matters of substantial relevance to the community-the code provides appropriate
community safeguards for the matters covered by the code; or
ii) in a case where the code does not deal with matters of substantial relevance to the community-the code deals with
the matters covered by the code in an appropriate manner; and
e) the ACMA is satisfied that, before giving the copy of the code to the ACMA:
i) the body or association published a draft of the code and invited participants in that section of the industry to make
submissions to the body or association about the draft within a specified period; and
ii) the body or association gave consideration to any submissions that were received from participants in that section of
the industry within that period; and
f) the ACMA is satisfied that, before giving the copy of the code to the ACMA:
i) the body or association published a draft of the code and invited members of the public to make submissions to the
body or association about the draft within a specified period; and
ii) the body or association gave consideration to any submissions that were received from members of the public within
that period; and
g) the ACMA is satisfied that the ACCC has been consulted about the development of the code; and
h) the ACMA is satisfied that the Telecommunications Industry Ombudsman has been consulted about the development of the code; and
i) the ACMA is satisfied that at least one body or association that represents the interests of consumers has been consulted
about the development of the code; and
j) in a case where the code deals with a matter set out in paragraph 113(3)(f)-the ACMA is satisfied that the Privacy
Commissioner has been consulted by the body or association about the development of the code before the body or
association gave the copy of the code to the ACMA; and
k) the ACMA has consulted the Privacy Commissioner about the code and consequently believes that he or she is satisfied
with the code, if the code deals directly or indirectly with a matter dealt with by:
i) the National Privacy Principles (as defined in the Privacy Act 1988 ); or
ii) other provisions of that Act that relate to those Principles; or
iii) an approved privacy code (as defined in that Act) that binds a participant in that section of the telecommunications
industry or the e-marketing industry; or
iv) provisions of that Act that relate to the approved privacy code.
2) The ACMA must register the code by including it in the Register of industry codes kept under section 136.
3) A period specified under subparagraph 1) e) i) or 1) f) i) must run for at least 30 days.
4) If:
a) an industry code (the new code ) is registered under this Part; and
b) the new code is expressed to replace another industry code;
the other code ceases to be registered under this Part when the new code is registered.
Note: An industry code also ceases to be registered when it is removed from the Register of industry codes under section 122A. http://www.austlii.edu.au/au/
legis/cth/consol_act/ta1997214/s117.html
C HAPTER 7
121
enues for the government, giving it substantial clout in policy
councils. Again, it may resist any attempts to further regulate
it.
Meanwhile, there are costs associated with any new administrative mechanism, even one as simple as the development,
registration, and updating of a code of conduct. Potential costs
should be factored into the cost-benefit analysis when considering adopting such a regime.
Additionally, adding intelligence to the middle of the network, and encouraging gatekeepers to use this intelligence, is
sub-optimal from a network design perspective. Like regulators
in developing countries, ISPs may themselves face resource
constraints to enforce their code. ISPs may or may not see sufficient incentives to do so. ISPs often have to balance multiple
interest and desires regarding spam, including:
• A desire to attract and retain bad-acting but paying customers,
• A desire to avoid the cost of transmitting spam through
their networks, and
• A desire to avoid the regulatory risks and costs of transmitting spam.
ISPs may over-enforce the provisions of their own code,
resulting in messages not getting delivered to recipients. This
would be a far worse outcome, many argue, than dealing with
the current deluge of spam. ISPs may also not be as sensitive to
the rights of free expression, and most speech protections do
not extend to non-state actions, often allowing private actors to
block otherwise protected speech.
Meanwhile, ISPs would likely pass anti-spam costs along
to end users, perpetuating the already-vicious cycle of spammers making the rest of the internet’s users pay for their bad
acts. In a developing country context, high internet access costs
are already a major barrier to widespread ICT adoption. Cost
concerns, however, should be seen in the context of the spam
problem itself, which is adding to the cost of internet access
and helping criminals to perpetrate fraud and disseminate
destructive viruses. These network ills are bad not only for
consumers, but for ISPs themselves.
Any legal and regulatory approach should seek to mitigate
these drawbacks. On balance, however, many jurisdictions will
likely find enforceable codes of conduct to be a sound policy
choice, because they distribute part of the enforcement burden
to stakeholders closest to the source of the spam problem – the
ISPs and the end users.
7.4 Education and Awareness
The ideal solution to spam would involve no new law
whatsoever. If consumers and businesses could take spam
fighting into their own hands, the problem would be solved at
the lowest cost and at the quickest rate. The brunt of anti-spam
enforcement would be borne at the furthest edges of the network and in the most distributed manner possible. Those who
pay the true costs of spamming – the end users – would ideally
take the lead in combating spam, while regulators focused their
enforcement resources on the largest, most complex cases.
122
Regulators would still have an important role to play, however, in educating consumers, businesses and ISPs about the
dangers of spam and the steps they can take to protect themselves against it.56 The London Action Plan includes some suggestions:
•
Regulators should develop a plan for consumer and ISP
education, posting information on their websites and
developing print materials for distribution to cybercafé
owners, consumers, businesses and ISPs.
•
Regulators should provide a simple method for consumers to make complaints about spam.
•
Regulators should create a special “combating spam” page
on their websites, providing information about anti-spam
practices and products. The web page should host practical advice on spam filters, warnings about phishing
attempts, viruses and scams carried out using e-mail and
other important tips for consumers. Examples of websites
in use today include:
–
Industry Canada’s page on “Recommended Best
Practices for Internet Service Providers and Other
Network
Operators”:
http://e-com.ic.gc.ca/epic/
internet/inecic-ceac.nsf/en/gv00329e.html
–
Recommendations of the Commission Nationale
de l’Informatique et des Libertés in France (CNIL
République
Française):
http://www.cnil.fr/index.
php?id=1539
–
Guidance provided by the Korea Spam Response
Centre of the Korea Information Security Agency, an
affiliated agency of the Ministry of Information and
Communication: http://www.spamcop.or.kr/eng/m_3_
2.html
–
The United States’ Federal Trade Commission’s
spam education pages: http://www.ftc.gov/bcp/
c o n l i n e / p u b s / b u s p u b s / s e c u r e y o u r s e r v e r. h t m
http://www.ftc.gov/bcp/conline/edcams/spam/
secureyourserver/index.htm
•
Regulators should also consider their ability to play a
central role in coordinating the sharing of best practices
among ISPs, especially in contexts where political will or
resources do not exist for the regulator to take an enforcement role. The regulator can also help educate ISPs about
some relatively simple technical measures. Specific measures include the latest information related to the blocking
of open relays,57 focus on “botnets,”58 and slowdowns of
traffic on port 25 that might make an enormous difference, particularly in developing countries.
Consumer and ISP education is a necessary component
of spam-fighting strategies, but efforts in this field have had
little effectiveness to date. This is not due to any fault in the
outreach techniques themselves, but rather due to the limited
vigour with which they have been pursued. It is challenging to
communicate technical information to a lay audience. Moreover, education efforts cannot succeed in isolation, without other
effective technological and regulatory measures. Substantially
greater efforts in this area are warranted and would pay large
dividends.
C HAPTER 7
7.5 Conclusion
Despite the challenges that are bound to lie ahead, regulators should encourage the adoption of an anti-spam law that is
harmonized, as much as possible, with those of other countries.
Such an anti-spam law might involve creating an enforceable
code of conduct for ISPs, placing the responsibility for mitigating spam closer to where the technical expertise lies. The
problem with anti-spam laws enacted to date is that they have
failed to create an enforceable regime or to bridge the divide
between governments and the technologists who have the real
expertise to solve the problem. While it is an imperfect remedy,
1
2
3
4
5
6
7
8
9
10
11
12
an enforceable code of conduct could help to erase the shortcomings of earlier anti-spam laws.
The effort to fight spam is not going to succeed through
pursuit of any one, single strategy. Success will be based on
international cooperation and a range of shared strategies,
including legal and regulatory mechanisms, technical improvements, market forces, and consumer-oriented solutions. The
development of ISP codes of conduct, and their enforcement
by regulators, can help stem the tide of spam and materially
reduce spam’s costs to ISPs and consumers.
Despite passage of many dozens of anti-spam statutes in jurisdictions across the globe, the problem has continued to worsen. See, e.g., David E. Sorkin, “Spam Legislation in the United States,” The John Marshall Journal of Computer and Information Law, Volume XXII, Number 1, at 4 (2003) (“…it is generally agreed that legislation has
failed to solve the spam problem.”) See also, Matthew Prince, “How to Craft an Effective Anti-Spam Law,” WSIS Thematic Meeting on Countering Spam, July 2004,
ITU Discussion Paper, at 10, at http://www.itu.int/osg/spu/spam/contributions/Background%20Paper_How%20to%20craft%20and%20effective%20anti-spam%20law.
pdf (“Few people would dispute that around the world the first generation of anti-spam laws has been an unqualified failure.”).
http://www.itu.int/ITU-D/treg/Events/Seminars/2004/GSR04/index.html
Business Software Alliance, 1 in 5 British Consumers Buy Software from Spam, Dec. 9, 2004, at http://www.bsa.org/uk/press/newsreleases/online-shopping-tips.cfm.
For instance, e-mail security provider IronPort Systems asserts that 72 per cent of e-mail sent is spam. See http://www.ironport.com/company/pp_sci-tech_today_0810-2005.html.
http://frwebgate.access.gpo.gov/cgi-bin/getdoc.cgi?dbname=108_cong_public_laws&docid=f:publ187.108.pdf
See Matthew Prince, “How to Craft an Effective Anti-Spam Law,” supra note 1, at 3.
For the most comprehensive resource on the world s anti-spam laws, see Christina Bueti, “ITU s Survey on Anti-Spam Legislation Worldwide,” July, 2005, at http://
www.itu.int/osg/spu/cybersecurity/docs/Background_Paper_ITU_Bueti_Survey.pdf.
AOL claims that spam is down 85 per cent from two years ago, based upon consumer complaint information. However, such a claim does not account for the effectiveness that their filters may have achieved on behalf of customers, nor the changing perceptions of consumer about how much spam is acceptable. The same
article that reported AOL s claim of less spam concludes, “But statistics show that the amount of spam is still huge – even worse than it was when the federal act [the
CAN-SPAM Act of 2003] was introduced two years ago.” See http://www.crmbuyer.com. See also http://www.washingtonpost.com/ (27 December 2004). There is a
dearth of reliable industry-wide data, which is not surprising in light of the distributed nature of the problem and the competition between ISPs to provide the best
anti-spam services to consumers.
For a review of some of the many recent spam statistics, see Bueti, “ITU s Survey on Anti-Spam Legislation Worldwide,” supra note 5; see Michael Geist, “Untouchable: A Canadian Perspective on the Anti-Spam Battle,” June, 2004, at 2, at http://www.michaelgeist.ca/geistspam.pdf; see also, Derek Bambauer, John Palfrey, and
David Abrams, “A Comparative Analysis of Spam Laws: the Quest for Model Law,” June 2005, at 7 – 8, at http://www.itu.int/osg/spu/cybersecurity/docs/Background_Paper_Comparative_Analysis_of_Spam_Laws.pdf.
“Phishing” refers to a scam in which perpetrators send an email purporting to be from a legitimate business (such as a bank) and ask recipients to provide personal
(often financial) information. Victims believe they are complying with a bona fide request, when they are being tricked into providing information to thieves. “Pharming” refers to a scheme in which victims clicking on a website are unknowingly diverted to a duplicate or fake website, where they can be fleeced.
See Chairman s Report, ITU WSIS Thematic Meeting on Cybersecurity, June – July, 2005, p. 2, point 12, at http://www.itu.int/osg/spu/cybersecurity/chairmansreport.
pdf (citing a speech by Spamhaus CEO Steve Linford).
http://news.bbc.co.uk/2/hi/business/3426367.stm.
C HAPTER 7
123
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
The AOL legal department posts decisions and litigation to their website at http://legal.web.aol.com/decisions/dljunk/. See also http://www.theregister.co.uk/2005/08/10/
aol_spam_sweepstake/ (regarding the AOL gold bars raffle, in which they planned to give away the assets seized from a major spammer).
See http://abcnews.go.com/Technology/PCWorld/story?id=1029922&ad=true.
This discussion paper uses the term “regulators” in the broad sense to include any governmental entity that has been given the mandate to combat spam. Thus, the
term “regulators” for this chapter may mean national telecommunications or ICT regulatory authorities, consumer protection authorities or data protection administrations.
See David R. Johnson, Susan P. Crawford, and John G. Palfrey, Jr., The Accountable Net: Peer Production of Internet Governance, 9 VA. J. L. & TECH. 9 (2004).
BBC, supra note 10.
The four modes of Internet regulation were popularized in Lawrence Lessig s ground-breaking book, Code and Other Laws of Cyberspace, in 1999 (New York: Basic
Books).
See http://www.itu.int/osg/spu/spam/background.html and, in particular, the Chairman s Report, at http://www.itu.int/osg/spu/spam/chairman-report.pdf.
Ibid., point 24 at 4.
It should be noted that even the United States Federal Trade Commission, which is a relatively well-funded regulatory body, had only brought “over 70 cases” as of
July, 2005. In light of the billions of spam messages per day, the notion that such an enforcement effort is unlikely to have much effect undoubtedly is apparent to
many governments choosing whether or not to devote resources to fighting spam locally. Ibid., point 19, at 3.
See Suresh Ramasubramanian, “OECD Task Force on Spam Report: Spam Issues in Developing Countries,” May, 2005, at http://www.oecd.org/dataoecd/5/47/34935342.
pdf.
See http://www.itu.int/ITU-D/treg/related-links/links-docs/Spam.html for a list of voluntary and enforceable ISP codes of conduct.
See generally Jonathan Zittrain, “Internet Points of Control,” 43 Boston College Law Review 653 (2003). See also, J.H. Saltzer, D.P. Reed, and D.D. Clark, “The
End-to-End Argument in Systems Design,” at http://www.reed.com/Papers/EndtoEnd.html and “The End of the End-to-End Argument” at http://www.reed.com/
dprframeweb/dprframe.asp?section=paper&fn=endofendtoend.html (“But in many areas of the Internet, new chokepoints are being deployed so that anything new
not explicitly permitted in advance is systematically blocked.”)
See John Spence, “Pennsylvania and Pornography: CDT v. Pappert Offers a New Approach to Criminal Liability Online,” 23 J. Marshall J. Computer & Info. L. 411
(Winter, 2005) (a good general discussion of the role of ISPs in the network and the difficulties they face).
http://www.maawg.org/about/roster/
Many technical working groups have focused on anti-spam-related standards, technologies, and best practices. The IETF, ISOC, and other groups have supported
efforts that have involved representatives of ISPs, including the now-scuttled MARID Project (see http://www.internetnews.com/bus-news/article.php/3407431),
which was preceded by the Anti-Spam Research Group (at http://asrg.sp.am/).
See Renai LeMay, Gmail Tries Out Antiphishing Tools, CNET NEWS.COM, Apr. 4, 2005, at http://news.com.com/Gmail+tries+out+antiphishing+tools/2100-1029_35653794.html.
See Anick Jesdanun, Battle Against Spam Shifts to Containment, ASSOCIATED PRESS, Apr. 15, 2005, at http://finance.lycos.com/qc/news/story.aspx?story=48398343.
Consider the remarks of Randall Boe, executive vice president of AOL, when he said that “Spam has become the single largest customer problem on the Internet.”
(Quoted in Thomas Claburn, “Four Big ISPs Sue Hundreds of Spammers,” 10 March 2004, Information Week, at http://www.informationweek.com/).
As one illustration of the fact that spam can be traced, see http://www.channelregister.co.uk/2005/09/20/spam_map/.
Consider, for instance, that MAAWG is already promoting industry-wide codes of conduct. See http://www.maawg.org/about/.
Bambauer, Palfrey, and Abrams, “A Comparative Analysis of Spam Laws: the Quest for Model Law,” supra note 9, at 11.
Prince, “How to Craft an Effective Anti-Spam Law,” supra note 1, at 4.
Ibid., at 6. Mr. Prince argues: “The most effective anti-spam laws are action laws that focus on the problems prosecutors face and work to resolve them. If we want
anti-spam laws to be effective, our job must be to identify the costs faced by prosecutors and craft laws to reduce those costs.”
Accessible online at http://www.uncitral.org/pdf/english/texts/electcom/05-89450_Ebook.pdf.
See http://www.itu.int/osg/spu/spam/ for a catalogue of existing anti-spam laws on the books in jurisdictions around the world.
Many analysts predicted the failure of these laws at the time they were passed. For one example of a United States-based consultancy, consider Gartner s report,
Maurene Caplan Grey, Lydia Leong, Arabella Hallawell, Ant Allan, and Adam Sarner, “Spam Will Likely Worsen Despite US Law,” 3 December 2003, at http://www.
gartner.com/resources/118700/118762/118762.pdf.
See BBC News, “US Still Leads Global Spam List,” 7 April 2005, at http://news.bbc.co.uk/1/hi/technology/4420161.stm (citing a study by security firm Sophos that
the US is responsible for sourcing 35 per cent of the world s spam).
See the FAQ page for the Coalition Against Unsolicited Commercial Email, at http://www.cauce.org/about/faq.shtml#offshore.
One interesting, as-yet-theoretical variant to the state-focused enforcement mechanism is the “bounty hunter” system proposed by Prof. Lawrence Lessig of Stanford
Law School. Prof. Lessig has “bet [his] job” on the notion that such a distributed system, established by law but pushing out enforcement authority to netizens, would
work if enacted. See http://www.lessig.org/blog/archives/000787.shtml.
The Australian law, which took effect in 2003, can be found online (in an unofficial version) at http://scaleplus.law.gov.au/html/pasteact/3/3628/0/PA000260.htm
For example, the text of the Communications Decency Act Section 230 in the United States provides immunity to the providers of “interactive computer services”
for the content published on their network. These providers are defined as follows: “The term `interactive computer service means any information service, system,
or access software provider that provides or enables computer access by multiple users to a computer server, including specifically a service or system that provides
access to the Internet and such systems operated or services offered by libraries or educational institutions.” http://www.fcc.gov/Reports/tcom1996.txt. By contrast,
the term “Internet access service” in the CAN-SPAM Act of 2003, as stated in the Telecommunications Act of 1934, as amended, reads: “The term Internet access
service means a service that enables users to access content, information, electronic mail, or other services offered over the Internet, and may also include access to
proprietary content, information, and other services as part of a package of services offered to consumers. Such term does not include telecommunications services.”
http://www4.law.cornell.edu/uscode/html/uscode47/usc_sec_47_00000231----000-.html.
Geist, “Untouchable,” supra note 8, at 17 (for a discussion of civil and criminal sanctions common in anti-spam legislation).
For discussion of the effectiveness of such a measure, see Prince, “How to Craft an Effective Anti-Spam Law,” supra note 1, at 9.
http://www.icpen.org/.
For discussion of the effectiveness of the state of Washington s use of such a measure in the United States, see Prince, “How to Craft an Effective Anti-Spam Law,”
supra note 1, at 6 and 10.
124
C HAPTER 7
48
49
50
51
52
53
54
55
56
57
57
For the full text of the Australian Telecommunications Act of 1997 that contains such provisions, see http://www.austlii.edu.au/au/legis/cth/consol_act/ta1997214/s117.
html et seq.
The Australian Direct Marketing Association (ADMA) has also established a Code of Conduct. Where such an organization exists, such a code is another logical,
parallel step. Many countries will not have such an entity in place, in which event a legal provision mandating a parallel process of this sort would not make sense.
Consider the findings of the New Zealand regulators with respect to the most effective mode of enforcement: “A civil penalty regime where the emphasis is on
ISPs/carriers taking action in response to customer complaints is considered to be the best approach. This is because most spam in New Zealand originates from
overseas and the ISP/carrier will often best be placed to put in place the appropriate technical measures to deal with it. In addition, if spam is originating from an
address/number hosted by another ISP/carrier in New Zealand, then the user s ISP/carrier can approach the sender s ISP/carrier and seek action by that ISP/carrier
against the sender. If complaints cannot be satisfactorily resolved in this way then the user s ISP/carrier can forward the matter on to the enforcement agency to
consider whether an investigation or further action is appropriate.” Ministry of Economic Development (NZ), “Legislating against Unsolicited Electronic Messages
Sent for Marketing or Promotional Purposes (Spam) – Enforcement Issues – Cabinet Paper,” at http://www.med.govt.nz/pbt/infotech/spam/cabinet/paper-two/papertwo-03.html#P31_3192.
G-REX is an online discussion platform reserved for policy-makers and regulators> For more information, see: http://www.itu.int/ITU-D/grex/index.html.
See http://www.truste.org/.
The process under way at the Messaging Anti-Abuse Working Group may well provide extremely useful guidance on this front, both as a matter of process and of
substance. See http://www.maawg.org/news/maawg050711.
See http://www.ftc.gov/os/2004/10/041012londonactionplan.pdf. See also, for particular suggestions, http://www.ftc.gov/bcp/conline/edcams/spam/zombie/index.htm.
For a letter sent to 3,000 ISPs, as part of this initiative, see http://www.ftc.gov/bcp/conline/edcams/spam/zombie/letter_english.htm.
The specific suggestions for such zombie-prevention measures will vary over time. Some initial recommendations, derived as part of the London Action Plan meeting
and related efforts, include: 1) blocking port 25 except for the outbound SMTP requirements of users authenticated by the ISP to run mail servers designed for client
traffic and other carefully accredited purposes; 2) exploring implementation of Authenticated SMTP on port 587 for clients who must operate outgoing mail servers;
3) applying rate-limiting controls for email relays; 4) identifying computers that are sending atypical amounts of email, and take steps to determine if the computer
is acting as a spam zombie. When necessary, quarantining the affected computer until the source of the problem is removed; 5) providing, or pointing customers to,
easy-to-use tools to remove zombie code if their computers have been infected, and provide the appropriate assistance; and, 6) the shutdown of open relay servers
after appropriate notice and inquiry. Regarding the first of these suggestions, related to port 25, Industry Canada (in a separate context), recommends, “ISPs and
other network operators should limit, by default, the use of port 25 by end-users. If necessary, the ability to send or receive mail over port 25 should be restricted to
hosts on the provider s network. Use of port 25 by end-users should be permitted on an as-needed basis, or as set out in the provider s end-user agreement / terms
of service.” http://e-com.ic.gc.ca/epic/internet/inecic-ceac.nsf/en/gv00329e.html.
The New Zealand regulators note: “The enforcement agency would be seen as also having a role in educating users/consumers on how to deal with spam in conjunction with the industry as well as a role in educating business and other organisations on how to comply with the legislation along with the Ministry of Economic
Development, which will be responsible for administering the legislation, and organisations such as the Direct Marketing Association.” Ministry of Economic
Development (NZ), “Legislating against Unsolicited Electronic Messages Sent for Marketing or Promotional Purposes (Spam) – Enforcement Issues – Cabinet
Paper,” supra note 47, at http://www.med.govt.nz/pbt/infotech/spam/cabinet/paper-two/paper-two-03.html#P31_3192.
For a description of open mail relays and their importance to the spam issue, see http://en.wikipedia.org/wiki/Open_mail_relay.
For a definition of botnet, see http://en.wiktionary.org/wiki/botnet.
C HAPTER 7
125
8 MAKING BROADBAND WORK FOR ALL
Regulators and policy-makers view the advent of broadband networks and services as both a challenge and an opportunity. What, really, is “broadband”? To the layman, it often
translates as “fast internet access.” But as the edition of Trends
has illustrated, the broadband revolution is a multi-layered
phenomenon, with technological, economic and social aspects
that will become more apparent as this decade wears on. One
thing is clear: broadband is not a passing phase or a high-end
niche market. It is literally the future of telecommunications.
Policy-makers and regulators around the world will have to
come to grips with it and learn to exploit its opportunities – in
other words, learn how to make broadband work for all.
8.1 What Is Broadband?
In the most basic definition, broadband refers to an array
of digital, packet-switched network technologies that allow the
transport of digital bits at high speeds. These technologies are
both wireless and wire-line, and they include both upgrades to
existing networks (for example, xDSL or 2.5G networks) and
entirely new infrastructure (such as all-fibre networks, WLANs
and 3G systems). Generally, networks with bandwidth capacities of 256 kbit/s or more can be termed “broadband,” although
that threshold may well shift higher as new technologies push
the envelope on throughput.
What may be more important than network capacity is
what one can do with broadband networks. Instead of the old,
single-purpose networks, broadband networks can carry any
combination of voice, data and multimedia (graphics, video and
audio), in any format. Indeed, broadband networks are already
generating new permutations on old media: audio “podcasts”
downloaded to portable players from websites, chat functions
incorporated into online video games. The list goes on. The
evolution of new applications is suddenly without boundaries,
and human ingenuity is now free to pursue services and applications that will improve lives and bolster economies.
The obvious conclusion is that the term “broadband” does
not just mean an interesting set of network technologies. It is
an entirely new paradigm, potentially as different from standard voice telephony as telephony is from the telegraph services
of 150 years ago. Never before has there been such power to
combine images and information in ways that can actually aug-
C HAPTER 8
ment the user’s experience into something more enriched than
actually being there.
8.2 Why Should I Care about Broadband?
Broadband networks amplify the internet, certainly, but
they go much further than that. A doctor 10,000 kilometres
away can check the progress of a patient’s care using a remote
live-camera feed, with an in-screen box showing the patient’s
records to facilitate diagnosis and treatment. A master of traditional dance in one country can teach students the intricacies
of an ancient dance step to émigrés in a hundred other countries at the same time, ensuring its survival for generations to
come. Such moments are as ephemeral as the flap of a butterfly’s wing – and just as priceless. Now imagine such moments
taking place literally millions of times each day across disciplines ranging from dance to dentistry to demographics.
The single most important thing to absorb about broadband technologies is that they drive intelligence and ingenuity
to the edge of networks. More than ever before in the history
of telecommunications, it will be not so much the network
but rather the people connected to it that count. ICT technology may never catch up to human creativity and diversity, but
broadband networks will allow it to remain closely linked. The
power of computing to generate and organize knowledge – or
to germinate and nurture art – will suffocate without the media
to convey it from one person to another. Broadband networks
empower individuals and groups to create and collate, innovate
and inspire, without restrictions of time and distance.
As they empower individuals, broadband capabilities will
increase the potential for generating content that will be relevant, meaningful and understandable to communities. The
key to sustainable network services is demand. And the key
to demand is providing useful, culturally sustaining content,
in local languages, about local circumstances as well as global
realities. Although it certainly will not happen overnight, there
is no reason why individuals in the remotest areas cannot eventually become broadcasters in their own communities, educators in their own homes, and performing artists for worldwide
audiences.
127
8.3 How Can I Get Broadband?
For increasing numbers of consumers, the answer is that
they already have it – and are likely to get more of it. In many
countries, broadband is now available in several different user
niches: at home on a desktop PC or with a Wi-Fi-equipped
laptop in the airport. For these lucky users, the future will be
about convergence onto multiple platforms – interactive digital
televisions, broadband mobile phones and streaming video on
computers, just to name a few – and inter-modal competition.
For developing countries, the key to the broadband future
is the flexibility of the technologies coupled with the declining
costs of the network topologies. Broadband technologies can
increasingly be either fixed or mobile, and they can convey any
mélange of voice, data and multimedia content – all at a lower
marginal cost than earlier-generation, circuit-switched telecommunications networks. Moreover, advances in infrastructure
– particularly with wireless networking standards in the Wi-Fi
and WiMAX families – will allow more and more broadband
capabilities at lower cost.
In the context of developing markets, broadband will
clearly have to dovetail with the prevalent mobile flavour of
the telecommunication sectors there. For one thing, the age
of the ubiquitous wire-line network may never arrive in many
countries. And certainly, the broadband revolution will not be
obliged to wait around for it. Access networks are likely to continue to be mobile and wireless – including broadband wireless access (BWA) technologies. Mobile and fixed wireless will
likely converge, bringing a broadband capability to markets that
are essentially mobile, and will continue to be. For transport
and backhaul, a combination of network types will have to be
employed. This will undoubtedly include terrestrial wireless
links, some satellite hops and lit fibre.
Non-traditional suppliers will have to be part of the mix
in developing countries. For backhaul, this may mean infrastructure sharing between the fibre networks of universities
and the transport and energy sectors as well as leased fibre from
the private sector. In the access network sector, the decentralized nature of broadband networks will enlist smaller-scale,
local and regional operators – community groups, universities, municipalities and entrepreneurs – to set up “hotspots”
or wide-area networks that can be linked back to larger operators’ networks. The network can be built from the periphery
inwards, by local operations providing local content and services, generating demand from the ground up.
8.4 What Can Regulators Do?
Regulators and policy-makers around the world are
reviewing their laws and regulations to judge whether they
provide a proper environment to help speed the opportunities
and benefits of broadband networks and the new services and
applications that ride on them. At this juncture, responses have
been varied – everything from initiating an overhaul of licensing and market-entry policies to doggedly restricting access
to VoIP services. As with other aspects of sector liberalization,
many governments feel the instinct to protect incumbents and
ensure that they take the lead in broadband investment.
128
It may do little good, however, to liberalize traditional
telephony markets, or even mobile service markets, while protecting incumbents from inter-modal or broadband competition. Growing numbers of countries are adopting regulations
that allow for open market entry at all levels and layers, including applications and services such as VoIP and internet access.
Of course, in markets undergoing a transition to competition,
it is important for regulators to work towards a level playing
field for providers of IP-based services. This will often mean
taking steps to ensure that network operators interconnect and
provide open access to support infrastructure and some network capabilities.
In many areas of regulation, it may be possible to “regulate down” rather than “regulate up” to achieve competition.
That is, as competition becomes viable, and market forces
begin to discipline operators’ behaviour, it may be wiser to
reduce the regulatory burden on all operators – incumbents
and new entrants alike – rather than imposing the same regulatory structure on competitors that has always been applied to
the monopoly incumbent. Regulators must strike a balance
between giving incumbents too much latitude to obstruct
competition, on the one hand, and so many restrictions that it
stifles broadband investment, on the other.
One key area of regulatory practice, of course, is spectrum
management, and the role of spectrum management takes on
greater importance with the rise of broadband wireless access
(BWA) technologies. At the same time that spectrum-hungry
BWA equipment is being developed and deployed, other
advances, such as in the computing power of processors in
radio equipment, are causing more ferment in the spectrum
community than perhaps at any time in its history. The traditional paradigm of spectrum planning, allocation, assignment
and monitoring reflect largely the technological limits of an
earlier era. New radio technologies may free regulators from
some legacy spectrum management practices that may become
obsolete, or worse, too rigid for a wireless broadband world.
Technological developments such as spectrum-hopping, adaptive and directional antennae use and other techniques promise
to shift interference management from governments to operators and even to end-user terminals to allow greater sharing
and reuse of existing spectrum.
Spectrum regulators have begun to respond by granting
licensees with more flexibile use of technologies so that operators can deliver the services the market demands. This technology and service “neutrality” reinforces the overall trends
of decentralizing control over networks. This will allow the
application of BWA technologies, for example, to be more agile
in responding to local market realities. Indeed, in developing
countries generally, and in remote areas within developing
countries, in particular, where spectrum scarcity is far less an
issue than in developed countries, there is no reason why spectrum management policies could not be tailored to local realities, which may include a less dense environment of spectrum
use, allowing greater power and range for wireless systems that
do operate there.
In the end, perhaps the best analogy of the future broadband wireless environment is the highway analogy. Anyone
C HAPTER 8
can get on the highway, as long as they obey the rules of the
road. That is, as long as operators obey rules designed to limit
harmful interference, entry barriers can be lowered to allow
as many operators as demand warrants. In congested areas, of
course, those rules may need to be more stringent – as speed
limits are on urban highways. In the end, however, it makes
little sense to restrict market entry when technical demands do
not require it. That would be akin to establishing a nationwide
limit of cars allowed on any highway anywhere in the country,
based solely on the potential for congestion at the single busiest spot, at rush hour, in the country’s capital city.
C HAPTER 8
In the final analysis, policy-makers and regulators need
to address the broadband revolution in three ways. First, they
need to accept the reality of its coming and embrace its potential. Second, they need to radically revise the way they understand telecommunications, to understand all the options they
might pursue in accommodating the new networks, services
and applications. Finally, and most importantly, they need to
plan, with the advice and input of their industries, academic
institutions and civil societies, how to make broadband work
for all.
129
REGULATORY TABLES *
Table 1
Countries with a separate Regulatory Authority ................................................................ 133
Table 2
Status of the main fixed-line operators ............................................................................ 169
Table 3
Level of competition ....................................................................................................... 203
* Regulatory profiles as well as contact details for policy-makers and regulators are available on the TREG website at: http://www.itu.int/ITU-D/treg.
This data is extracted from the ITU World Telecommunication Regulatory Database.
R EGULATORY T ABLES
131
1. Countries with a separate Regulatory Authority
WORLD
Country
Name of Authority
Year
Legal document that created
created
the authority
Financed by
Reports to
Is it a collegial body?
Afghanistan
Afghanistan Telecom Regulatory Authority (ATRA)
2003 Presidental Decree
Website: http://www.trb.gov.af
Sector Ministry
(annual report)
Regulatory fees: 100%
Yes: 5 Members
Albania
Telecommunication Regulatory Entity
1998 Law 8288,18.2.1998 on
Telecommunication
Regulatory Entity
Website: http://www.ert.gov.al
Legislature
Other: Council of
Ministers
Licence fees: 17.1%
Spectrum fees: 38.3%
Numbering fees: 0.1%
Yes: 5 Members
Financial income: 44.4%
Other: 0.1%
Algeria
Autorité de régulation de la poste et des télécommunications (ARPT)
2000 Loi n° 2000-03
Legislature
Website: http://www.arpt.dz
Award/auction of other licence:
17.7%
Yes: 7 Members
Other: 20.6% Various products,
research fees, ICT training and
standardization
1
Angola
Institut Angolais des Communications (INACOM)
1999 Décret nº 12/99 du 25 juin
Website: http://www.inacom.og.ao
Sector Ministry
(annual report)
Award/auction of mobile licence:
30%
2.0%
Spectrum fees: 65%
Yes: 5 Members
Other: 3%
1
2
2004 data
pre-2004 data
Source: ITU World Telecommunication Regulatory Database.
R EGULATORY T ABLES
133
WORLD
Country
Name of Authority
Year
created
the authority
2
Financed by
Reports to
Argentina
Comisión Nacional de Comunicaciones (CNC)
Website: http://www.cnc.gov.ar
1990 Decreto 1185 de fecha 22 de
junio de 1990
Spectrum fees: 55%
Yes: 8 Members
Contributions from operator
turnover: 42%
Other: 2.5%
Australia
Australian Communications and Media Authority (ACMA)
1997 Austel - Telecommunications
Act 1989. ACCC - Trade
Practices Act 1974. ACA ACA Act 1997. Australian
Communications and Media
Authority Act 2004
1
Legislature
Sector Ministry
(annual report)
Website: http://www.acma.gov.au
Licence fees
Government appropriation
Spectrum fees
Numbering fees
Yes: 7 Members
turnover
Other: ACMA and ACCC telco
regulatory costs are recovered
from carriers via carrier licence
fees, based on each carrier's %
of total carrier eligible revenue.
Costs of administering numbering
arrangments are recovered
through numbering charges.
'Once off' fees (eg. spectrum
auction) are paid directly to
general goverment consolidated
revenue. At the time of auction, a
recurrent annual tax is set for
each band in the spectrum.
Bidders who use the particular
band are taxed pro-rata for the
megahertz population in the band
that they use, i.e. their share of
the total band tax.
Austria
Telecom-Control-Commission (Telekom-Control-Kommission, TKK)
1997 Telecommunications Act 1997 Sector Ministry
(annual report)
Website: http://www.rtr.at
turnover: 70,5 %
Other: 29,5% Broadcasting
activities are financed by
broadcasting companies (Data
for 2003)
Yes: 3 Members
1
2
2004 data
pre-2004 data
134
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Bahamas
The Public Utilities Commission
Website: http://www.pucbahamas.gov.bs
1999 Public Utilities Commission
Act, 1993
Other: The PUC is
required 3 months
after the end of the
financial year to
transmit an Annual
Report to the
Governor General
and Prime Minister
Yes: 3 Members
Licence fees
Government appropriation: 19.9%
turnover: 45.9%
Bahrain
Telecommunications Regulatory Authority
Website: http://www.tra.org.bh
2002 The Telecommunications Law No reporting
requirements
Licence fees: 100%
Yes: 5 Members
Bangladesh
Bangladesh Telecommunication Regulatory Commission (BTRC)
2002 Bangladesh
Legislature
Telecommunication Act, 2001 Sector Ministry
(annual report)
Website: www.btrc.org.bd
Other: 100% Govt. of Bangladesh Yes: 5 Members
Barbados
Fair Trading Commission
Website: http://www.ftc.gov.bb
2001 Fair Trading Commission Act
2000-31
Sector Ministry
(annual report)
Other Ministry
Yes: 11 Members
turnover
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
135
WORLD
Country
Name of Authority
Year
created
the authority
2
Reports to
Financed by
Belgium
Belgian Institute for Postal Services and Telecommunications
1993 Act of 21 March 1991 on the
reform of some economic
public companies. Act of 17
January 2003 concerning the
statute of the Belgium Post
and Telecommunication
regulator
2
Legislature
Sector Ministry
(annual report)
Other Ministry
Website: http://www.bipt.be
2%
Licence fees: 7%
Spectrum fees: 81%
Numbering fees: 6%
Yes: 4 Members
Other: 4% Office of the
ombudsperson
Belize
Public Utilities Commission (PUC)
1988
2
Benin
Direction de la Politique des Postes et Télécommunications (interim)
2002 Ordonnance n° 2002- 003 du Sector Ministry
31 janvier 2002
(annual report)
Other: The Authority is not yet
operational.
Yes: 5 Members
Bhutan
Bhutan Communications Authority
2000 Bhutan Telecommuncations
Act 1999
2
Website: http://www.bca.gov.bt
Sector Ministry
(annual report)
Government appropriation: 100%
No: Director
Bolivia
Superintendencia de Telecomunicaciones
Website: http://www.sittel.gov.bo
1995 Ley 1600 (Ley SIRESE)
Spectrum fees: 100%
No:
Telecommunicaction
Superintendent
1
2
2004 data
pre-2004 data
136
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
2
Financed by
Reports to
Bosnia and Herzegovina
Communications Regulatory Agency (CRA)
Website: http://www.cra.ba
1999 Decision combining the
Other: The Council
competencies of the
of Ministers
Independent Media
Commission and the
Telecommunications
Regulatory Agency issued by
Office of the High
Representative (OHR) in 2001
Licence fees: 70%
Spectrum fees: 7%
Numbering fees: 23%
No: Chief Executive
Officer (CEO)
Botswana
Botswana Telecommunications Authority (BTA)
1996 Telecommunication Act 1996
(No 15 of 1996)
2
Website: http://www.bta.org.bw
Sector Ministry
(annual report)
Licence fees: 91%
Spectrum fees: 9%
Yes: 5 Members
Brazil
Agência Nacional de Telecomunicações do Brasil (Anatel)
1997 Law Nº 9.9.472 of July 16,
1997
No reporting
requirements
Website: http://www.anatel.gov.br
48%
8.6%
Spectrum fees: 0.9%
Regulatory fees: 25.2%
Fines/penalties: 0.62%
Yes: 5 Members
Other: 0.1% Management
services, homologation
1
Brunei Darussalam
Authority for Info-communications Technology Industry of Brunei Darussalam (AiTi)
Website: http://www.aiti.gov.bn
2003 Authority for infocommunications technology
industry of Brunei
Darussalam Order, 2001
Other: Report to
Minister of
Communications
Confidential
Yes: 5 Members
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
137
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Bulgaria
Communications Regulation Commission
2002 Telecommunications Law
1
Website: http://www.crc.bg
Other: To the
National Assembly,
the President of the
Republic, the
Council of
Ministers, the
Council for
Electronic Media.
57,6%
Licence fees: 1,6%
Spectrum fees: 22,7%
Numbering fees: 2,6%
Fines/penalties: 0,2%
Yes: 5 Members
turnover: 15,3%
Burkina Faso
Autorité Nationale de Régulation des Télécommunications
1998 Loi n° 051/98/AN du 04
décembre 1998
Sector Ministry
(annual report)
Website: http://www.artel.bf
Award/auction of other licence
Licence fees
Spectrum fees: 45%
Numbering fees: 15%
Fines/penalties: 1%
No: Director General
turnover: 1%
Other: 4% Approval and
authorization fees
2
Burundi
Agence de Régulation et de Contrôle des Télécommunications
1997 Décret Présidentiel n° 100/182 Sector Ministry
(annual report)
No:
Administrator/DirectorGeneral
1
2
2004 data
pre-2004 data
138
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
1
Reports to
Financed by
Cameroon
Agence de Régulation des Télécommunications
1998 Loi n° 98/014 du 14 juillet 1998 Other: Submits an Licence fees
annual public report Spectrum fees
on activities
Numbering fees
Regulatory fees
Fines/penalties
turnover
Canada
Canadian Radio-television and Telecommunications Commission (CRTC)
1976 Telecommunications Act and
CRTC Act
Sector Ministry
(annual report)
Website: http://www.crtc.gc.ca
Yes: 7 Members
Other: Telecom activities are
funded by fees. Cable activities
are funded by entities regulated
under the Broadcasting act.
Cape Verde
Instituto das Comunicações e das Técnologias de Informação (ICTI)
2004 Résolution nº 1/2004 du 19
janvier 2004
1
Website: http//www.icti.cv
Sector Ministry
(annual report)
Yes: 3 Members
Central African Rep.
Agence chargée de la Régulation des Télécommunications (ART)
1
Chad
Office Tchadien de Régulation des Télécoms (OTRT)
1998 Loi 009/PR/98 du 17 août
1998
Website: http://www.otrt.td
Sector Ministry
(annual report)
Regulatory fees: 34,5%
Other: Chairman of Contributions from operator
the Board of
turnover: 13,7%
Directors
Other: 0,5% Transfers, fixed
assets
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
139
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Chile
Subsecretaría de Telecomunicaciones
1977 Decreto Ley Nº 1.762
Website: http://www.subtel.cl
Legislature
Sector Ministry
(annual report)
Other Ministry
0.5%
Spectrum fees: 45%
No: Subsecretariat of
telecommunications
Other: 54% Public Treasury
Colombia
Comisión de Regulación de Telecomunicaciones (CRT)
1994 Ley 142 de 1994, Decreto
1130 de 1999
1
Legislature
Sector Ministry
(annual report)
Other Ministry
Website: http://www.crt.gov.co
turnover: 100%
Financial income
Yes: 3 Members
Congo (Dem. Rep.)
Autorité de Régulation de la Poste et des Télécommunications du Congo ( A.R.P.T.C.)
Website: http://www.arptc.cd/
2002 Loi n° 014/2002 du 16
octobre 2002
Other: President of Numbering fees
the Republic
Other: Regulatory fee
Yes: 7 Members
Costa Rica
Autoridad Reguladora de los Servicios Públicos (ARESEP)
1963 Ley 258 de 1941; Ley 3226
Legislature
de 1963: facultades
Other: Annual
regulatorias; Ley 7593 del
activities report
año 1996 se transformó en la
ARESEP
Website: http://www.aresep.go.cr
Other: 100% Regulatory fees
covered directly by the operator
Yes: 5 Members
1
2
2004 data
pre-2004 data
140
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
1
Reports to
Financed by
Côte d'Ivoire
Agence des télécommunications de Côte d’Ivoire (ATCI)
1995 Loi n° 95-526 du 07 juillet
1995 portant Code des
Télécommunications
Sector Ministry
(annual report)
Other: Ministry of
the Economy and
Finance
Website: http://www.atci.ci
8.74%
Yes: 10 Members
turnover: 11.96%
Other: 4.61% Various
1
Croatia
Croatian Institute of Telecommunications
2000 Law on Telecommunications
Website: http://www.telekom.hr
Other: Annual
report to the
Government
Spectrum fees: 50%
Numbering fees: 24%
Yes: 7 Members
turnover: 25%
Other: 1% technical examinations
and additional incomes
Cyprus
Office of the Commissioner of Elecronic Communications & Postal Regulation
Website: http://www.ocecpr.org.cy
2002 The Telecommunications and
Postal Services Regulation
Law of 2002
Other: Annual
Report to the Head
of State
Licence fees: 75.2%
Fines/penalties: 0.5 %
No: Commissioner of
Electronic
Communications and
Postal Regulation
Financial income: 3.4 %
Other: 0.02% Postal licence fees
Czech Republic
Czech Telecommunication Office
2000 Act on Electronic
Communications and on
Amendments to Other Acts
No. 127/2005 Coll.,
Website: http://www.ctu.cz
Other: Annual
Report is presented
to the Government
and the Parliament
Yes: 5 Members
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
141
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Denmark
National IT and Telecom Agency (NITA)
1991 Finance Act 2002
1
Website: http://www.itst.dk/
Sector Ministry
(annual report)
Other: 35% Various services
Other: Annual
Status Report.
Annual activity
account on
financial
circumstances and
performance.
No: Director-General
Dominican Rep.
INDOTEL
Website: http://www.indotel.org.do
1998 Ley General de
Telecomunicationces
No. 153-98
Other: Annual
report to the
Executive for
presentation to the
National Congress
of the Dominican
Republic
Yes: 5 Members
turnover
Other: In addition, revenues are
received corresponding to the
use of the radio spectrum public
domain; the fees established, as
appropriate, under the
procedures for the issue of
concessions and licences, in
accordance with the regulations;
from the returns generated by its
own capital resources; from
budgetary appropriations
assigned to it, as appropriate, by
the central Government; and from
any other possible sources.
1
2
2004 data
pre-2004 data
142
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Ecuador
Consejo Nacional de Telecomunicaciones (CONATEL)
1995 Ley Reformatoria a la Ley
No reporting
Especial de
requirements
Telecomunicaciones. R.O.
770 del 30 de agosto de 1995
Website: http://www.conatel.gov.ec
10%
10%
Licence fees: 10%
Spectrum fees: 70%
Yes: 7 Members
Egypt
National Telecommunication Regulatory Authority (NTRA)
1998 Presidential Decree
Sector Ministry
(annual report)
Other: NTRA Board
of Directors
Website: http://www.tra.gov.eg/
Licence fees
Spectrum fees
Regulatory fees
Fines/penalties
No: Executive
President
Financial income
El Salvador
SIGET
Website: http://www.siget.gob.sv
1996 Ley de creación de la SIGET
Legislature
1.54%
Licence fees: 5.54%
Yes: 3 Members
Other: 22.75% Accumulated
surpluses, electricity charges
and miscellaneous charges
Eritrea
Communications Department
Website: http://www.cd.gov.er
1998 The Communications
Proclamation No. 102/1998
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
143
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Estonia
Estonian National Communications Board
1998 Government of the Republic
Act
1
Other: 100% State Budget
Ethiopia
Ethiopian Telecommunications Agency
1996 Telecommunication
proclamation No. 49/1996
1
Website: http://www.sa.ee
Sector Ministry
(annual report)
Website: http://www.telecom.net.et/~eta/
Legislature
Sector Ministry
(annual report)
No: General Manager
Finland
Finnish Communications Regulatory Authority
1988 Act on Telecommunications
Administration
Website: http://www.ficora.fi
Other: Half-yearly
financial report to
the sector Ministry
with evaluation of
achievement of
targets
Licence fees: 18%
Spectrum fees: 22%
Numbering fees: 13%
Other: 47% refund for collection
of television fees 34%, domain
name fees 9%, other fees 4 %
France
Autorité de Régulation des Communications Electroniques et des Postes (ARCEP)
Website: http://www.art-telecom.fr/
1997 Loi de réglementation des
télécommunications n° 96659 du 26 juillet 1996
Legislature
Sector Ministry
(annual report)
Other: Annual
report to president
of France
Yes: 7 Members
1
2
2004 data
pre-2004 data
144
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Gabon
Agence de Régulation des Télécommunications (ARTEL)
2001 Loi N°005/2001
Sector Ministry
(annual report)
Other Ministry
Other: ARTEL also
reports to finance
ministry. It may also
report to audit
office for financial
matters
Website: http://www.artel.ga
Licence fees: 70%
Spectrum fees: 12%
Yes: 6 Members
turnover: 4%
The Administrative and Financial
Directorate of ARTEL.
Gambia
Public Utility Regulatory Authority (PURA)
2004 PURA Act 2001
Website: http://www.pura.gm/
Legislature
Other: Secretary of
State for Finance
and Economic
Affairs, who is
responsible for the
Administration of
the Act. He in turn
will submit it to the
National Assembly
Yes: 6 Members
Georgia
Georgian National Communications Commission
2000 Georgian Law on
Other: Annual
Telecommunications and Post Report to the
President of
Georgia
Website: http://www.gncc.ge
Yes: 3 Members
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
145
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Germany
Federal Network Agency for Electricity, Gas, Telecommunication, Post and Railway (Section 115)
Website: http://www.bundesnetzagentur.de
1998 Telecommunications Act of
25 July 1996
1
Legislature
Other: Annual
Report [to the
interested public;
the reporting to
Legislature is made
by the bi-annual
Activity Report (cf.
Sections 121 and
122 of the
Telecommunications
Act)].
Spectrum fees: 32%
Numbering fees: 13%
No: President
Other: 9.3% amateur radio, digital
signature, radio equipment and
telecommunications terminal
equipment, operator's certificate
in the aeronautical mobile service,
measurements for third parties
(i.e. environmental impact
measurements, Leeheim Satellite
Monitoring station, etc.) and post
(postal licenses)
Ghana
National Communications Authority
1997 National Communications
Authority Act 524 of 1996
Website: www.nca.org.gh (this is under construction)
Sector Ministry
(annual report)
Licence fees: 16%
Spectrum fees: 46%
Fines/penalties: 5%
Yes: 7 Members
Financial income: 2%
Greece
National Telecommunications and Post Commission, Greece (EETT)
1992 Law 2075/1992
Sector Ministry
(annual report)
Website: http://www.eett.gr
Licence fees: 0.48%
Yes: 9 Members
turnover: 12.34%
Other: 72.87% EETT's reserves
from previous years
1
2
2004 data
pre-2004 data
146
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
2
Reports to
Financed by
Grenada
National Telecommunications Regulatory Commission
Website: http://www.ectel.int/grd/
2001
2
Yes: 5 Members
Guatemala
Superintendencia de Telecomunicaciones
Website: http://www.sit.gob.gt/
1996 Ley General de
Telecomunicaciones (Decreto
94-96 del Congreso de la
República)
2
10%
No: Superintendent
Other: 90% Fines, 3%; bank
interest, 81%; administrative
charges, 6%
Guinea
Direction nationale des postes et télécommunications
1992
1
Sector Ministry
(annual report)
No: National Director
of Posts and
Telecommunications
Guinea-Bissau
Institut des Communications de la Guinée-Bissau (ICGB)
1999 Décret-loi nº 3/99, 20 août
2
Other: The Regulatory Authority
(DNPT) is financed out of the
Public Treasury.
Sector Ministry
(annual report)
Other Ministry
Website: http://www.icgb.org
Licence fees: 3%
Spectrum fees: 62%
Yes: 3 Members
Guyana
Public Utilities Commission
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
147
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Haiti
Conseil National des Telecommunications (CONATEL)
1969 Décret-loi de création du
Sector Ministry
CONATEL 27 Septembre 1969 (annual report)
Website: N/A
Licence fees
Spectrum fees
turnover
Other: Type approval
2
Honduras
Comisión Nacional de Telecomunicaciones
Website: http://www.conatel.hn
1996 Ley Marco del Sector de
Telecomunicaciones
Yes: 5 Members
3 principals, 2
substitutes
Hungary
National Communications Authority
Website: http://www.nhh.hu
1999 The present structure is
created by Act C of 2003 on
Electronic Communications
2
Legislature
Other: Annual
Report to the
Government
Licence fees: 1%
Spectrum fees: 56%
Numbering fees: 12%
Yes: 7 Members
turnover: 13%
Other: 18% V.A.T. reimbursed
Iceland
Post and Telecom Administration
Website: http://www.pta.is
1997 Law on the Post and Telecom
Administration, no. 147/1996
2%
Licence fees: 19%
Spectrum fees: 44%
No: Managing Director
Other: 35% Operating fees
India
Telecom Regulatory Authority of India
Website: http://www.trai.gov.in
1997 Telecom Regulatory Authority Legislature
of India Act-1997
Other: Annual
Report to the
Parliament
No: Chairperson
1
2
2004 data
pre-2004 data
148
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
1
Reports to
Financed by
Indonesia
Directorate General of Posts and Telecommunications
2003 Minister of Communications'
Decree No.31 regarding
Establishment Indonesian
Telecommunication
Regulatory Body
Sector Ministry
(annual report)
Website: http://www.postel.go.id
Yes: 5 Members
Iran (I.R.)
Communications Regulatory Authorithy
Website: http://www.cra.ir
2003 The law of duties and
Sector Ministry
powers of the Ministry of ICT, (annual report)
2003
Other: 100% At present from the
Government Fund. The following
possible revenues are
transferred to the Government
account: License fees, Spectrum
fees, Numbering fees,
Yes: 7 Members
Ireland
Commission for Communications Regulation (ComReg)
2002 Communications (Regulation)
Act 2002
2
Sector Ministry
(annual report)
Website: http://www.comreg.ie
Licence fees: 61%
Yes: 3 Members
turnover: 30%
Other: Cable and MMDS, Wireless
Telegraphy Fees etc 2%, Postal
Levy 4%
Italy
Autorità per le Garanzie nelle Comunicazioni (AGCOM)
1998 L. 31 luglio 1997, n. 249
Legislature
Website: http://www.agcom.it
Yes: 9 Members
turnover: 39%
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
149
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Jamaica
Office of Utilities Regulation
Website: http://www.our.org.jm
1995 Office of Utilities Regulation
Act
Other: Annual
report is submitted
to Legislative
Branch through the
Minister of
Development
Yes: 3 Members
Other: 6.63% License application
processing fees (6.43%) and
grants (0.2%)
Jordan
Telecommunications Regulatory Commission (TRC)
1995 Telecommunication Law
No. 13 of 1995 and its
amendments of 2002
1
Website: http://www.trc.gov.jo
Other: Report to the Licence fees: 82.377%
prime minister
Yes: 5 Members
Kenya
Communications Commission of Kenya
1999 Kenya Communications Act,
1998
Website: http://www.cck.go.ke
Sector Ministry
(annual report)
Yes: 11 Members
turnover: 17.7%
Korea (Rep.)
Korea Communications Commission
1997 Article 37~44 of the
Telecommunications Basic
Act
Website: http://www.kcc.go.kr
Legislature
Yes: 7 Members
1
2
2004 data
pre-2004 data
150
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Kyrgyzstan
State Communications Agency
Website: http://www.gas.gov.kg
1997 Decree of the President of
the Kyrgyz Republic 280, 7
October 1997
Other: Annual
report to the
Government
No: Director
turnover: 61.6%
Latvia
Public Utilities Commission
Website: http://www.sprk.gov.lv
2001 Law on Regulators of Public
Services
1
Legislature
Yes: 5 Members
Lesotho
Lesotho Telecommunications Authority
2000 Lesotho Telecommunications
Authority Act 2000
2
turnover: 100%
Website: http://www.lta.org.ls
Sector Ministry
(annual report)
Other Ministry
No: Chief Executive
Officer
Liechtenstein
Office for Communications
Website: http://www.ak.li
1999 Telekommunikationsgesetz
(TelG) vom 20. Juni 1996,
LGBl. 1996 Nr. 132, Art. 42a
(Telecommunications Law of
20 June 1996, Liechtenstein
Legal Gazette 1996 No. 132,
Art. 42a)
Sector Ministry
(annual report)
Spectrum fees
Numbering fees
Regulatory fees
No: Director
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
151
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Lithuania
Communications Regulatory Authority of the Republic of Lithuania
2000 Governmental Resolution
No. 617 on setting up NRA
Legislature
Other: Annual
report to the
Parliament and the
Government
Website: http://www.rrt.lt
Other: 100% Income for services Yes: 5 Members
provided and work performed
(administrative charges). The
mechanism of financing CRA is
through the State budget,
whereby all income from
administrative charges collected
for supervision of use of radio
frequencies, numbers, etc. is
deposited in the State budget and
then reallocated as needed to the
CRA
Luxembourg
Institut Luxembourgeois de Régulation
Website: www.ilr.lu
1997 Loi modifiée du 21 mars 1997 Other: Annual
sur les télécommunications
financial report to
minister, who
supervises institute
Spectrum fees
Numbering fees
Regulatory fees
Yes: 3 Members
turnover
Financial income
Madagascar
Office Malagasy d'Etudes et de Régulation des Télécommuniations (OMERT)
1997 Décret n° 97-1077 du 28 août Sector Ministry
1997
(annual report)
Website: http://www.omert.mg
Licence fees
turnover: 31.41%
Other: 1% Testing and
verification : 0,21%; Laboratory
fees : 0.02%, Other financial
products : 0.78%.
1
2
2004 data
pre-2004 data
152
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
2
Reports to
Financed by
Malawi
Malawi Communications Regulatory Authority (MACRA)
1998 Communications Act 1998
Sector Ministry
(annual report)
Website: http://www.macra.org.mw
Licence fees: 23%
Yes: 8 Members
turnover: 38.5%
Malaysia
Malaysian Communications and Multimedia Commission
1998 Malaysian Communications
and Multimedia Commission
Act 1998
1
Website: http://www.mcmc.gov.my
Other: Report to
Licence fees: 40%
Sector Minister,
Spectrum fees: 54%
Financial Report to Financial income: 6%
Sector Minister,
Annual Report and
Industry
Performance Report
Yes: 9 Members
Maldives
Telecommunications Authority of Maldives
2003
Website: www.tam.gov.mv
Sector Ministry
(annual report)
Licence fees: 5%
Yes: 5 Members
Mali
Comité de Régulation des Télécommunications (CRT)
1999 Ordonnance 99-043 du
30/9/99 (art.43). Décret
00-227 PRM du10 mai 2000
fixant les modalités de
fonctionnement du CRT
Sector Ministry
(annual report)
Website: http://mali-reforme-telecom.mcmtl.com
Spectrum fees: 51%
Yes: 3 Members
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
153
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Malta
Malta Communications Authority
Website: http://www.mca.org.mt
1997 The Telecommunications
(Regulation) Act as amended
by the Malta Communications
Authority Act.
1
Sector Ministry
(annual report)
Licence fees: 14%
Yes: 5 Members
Mauritania
Autorité de Régulation
Website: http://www.are.mr
1999 Loi 99 -019 du 11 juillet 1999
et Loi 2000 - 018 du 25
janvier 2001
Legislature
Spectrum fees: 26%
Other: The report is Numbering fees: 5%
annual and is
addressed to the
turnover: 68%
Government and
Parliament
Provisional budget for financial
year 2004
Yes: 5 Members
Mauritius
Information and Communication Technologies Authority
2002 Information and
Communication Technologies
Act
Website: http://www.icta.mu
Sector Ministry
(annual report)
Yes: 7 Members
Mexico
Comisión Federal de Telecomunicaciones (COFETEL)
1996 Decreto de creación de la
Comisión Federal de
Telecomunicaciones.
Ley Federal de
Telecomunicaciones.
Other: Annual
report
Website: http://www.cofetel.gob.mx
Yes: 4 Members
1
2
2004 data
pre-2004 data
154
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Moldova
National Regulatory Agency in Telecommunications and Informatics
2000 Government decision No. 843 Other: Report to
of 17.08.2000
Government
Website: http://www.anrti.md
Licence fees: 1.2%
Yes: 3 Members
Mongolia
Communications Regulatory Commission
2002 Communications law of 1995
amended in 2001
Website: http://www.crc.gov.mn
Other: Government Spectrum fees: 22.4%
of Mongolia
Yes: 7 Members
Other: 4.8% bank interest rate etc.
Morocco
Agence Nationale de Réglementation des Télécommunications (ANRT)
1997 Dahir n°1-97-162 du 7 août
1997 portant promulgation de
la loi 24-96 relative à la poste
et aux télécommunications
Other: Director
general produces
annual report,
submitted to prime
minister and
published in the
"Bulletin Officiel"
Website: http://www.anrt.net.ma
Spectrum fees: 45%
turnover: 53%
Other: 2% Various administrative
fees related to the declaration of
value-added services and
equipment approval.
No: Ag. DirectorGeneral
Mozambique
Instituto Nacional das Comunicações de Moçambique (INCM)
1992 Decree 22/92
Sector Ministry
(annual report)
Website: http://www.incm.gov.mz
Licence fees: 45%
Yes: 5 Members
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
155
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Namibia
Namibian Communications Commission
1992 Namibian Communications
Commission Act, 1992
Website: http://www.ncc.org.na
Sector Ministry
(annual report)
No: Chairman
Nepal
Nepal Telecommunications Authority
Website: http://www.nta.gov.np
1998 Telecommunication Act, 1997 Sector Ministry
(annual report)
Licence fees: 100%
Yes: 5 Members
Netherlands
OPTA
Website: http://www.opta.nl
1997 Wet OPTA
Sector Ministry
(annual report)
Yes: 3 Members
New Zealand
Commerce Commission
Website: http://www.comcom.govt.nz
2001 Commerce Act 1986,
Legislature
Telecommunications Act 2001 Sector Ministry
(annual report)
Yes: 3 Members
Other: 36% recovered from
applicants and parties
Nicaragua
TELCOR
Website: http://www.telcor.gob.ni
1995 Ley General de
Telecomunicaciones y
Servicios Postales (Ley 200)
Legislature
Other: To the
Executive
2.61%
Licence fees: 65.62%
No: Chief Executive
Director
Other: 1.16% Examination fee,
various services
1
2
2004 data
pre-2004 data
156
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
1
Reports to
Financed by
Niger
Autorité de Régulation Multisectorielle (ARM)
Website: http://niger.arm-niger.org/
2004
2
Nigeria
Nigerian Communications Commission
1992 NCC ACT No 75 of 1992
Website: http://ncc.gov.ng
Legislature
Sector Ministry
(annual report)
Licence fees: 68%
Yes: 9 Members
turnover: 12%
Other: 4.5% Type Approval
(3.5%), administrative fees (1%).
Norway
Norwegian Post and Telecommunications Authority
1987
Sector Ministry
(annual report)
Website: http://www.npt.no
Licence fees: 5%
Spectrum fees: 40%
Numbering fees: 4%
No: Director
Other: 3% Postal regulatory fees
Oman
Telecommunication Regulatory Authority
2002 Telecommunications
Regulatory Act 2002
Website: http://www.tra.gov.om
Other: Annual
reporting to the
6.6%
Council of Ministers Award/auction of other licence:
2.6%
Licence fees: 13.2%
Yes: 4 Members
turnover: 13.2%
Other: 0.003%
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
157
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Pakistan
Pakistan Teleccommunication Authority (PTA)
1996 Pakistan Telecommunication
Re-Organization Act 1996
Website: http://www.pta.gov.pk
No reporting
requirements
92.5%
Yes: 3 Members
1
Panama
Ente Regulador de los Servicios Públicos
1996 Ley N° 26 (29 de enero de
1996) por la cual se crea el
Ente Regulador de los
Servicios Públicos
2
Website: http://www.ersp.gob.pa
Other: Assembly of Contributions from operator
legislators
turnover: 1%
Other: 1% The fee charged,
in the case of the
telecommunication sector, to
service-providing companies for
control, monitoring and audit
services: 1% of annual gross
revenues for type B licensees
and 0.25% of annual gross
revenues for mobile cellular
telephone companies.
Papua New Guinea
PANGTEL
Website: http://www.pangtel.gov.pg/
1997
1
Yes: 3 Members
Licence fees
Paraguay
Comisión Nacional de Telecomunicaciones (CONATEL)
1995 Ley 642/1995 de
Telecomunicaciones
Legislature
Sector Ministry
(annual report)
Other: The
Executive
Website: http://www.conatel.gov.py
Yes: 5 Members
1
2
2004 data
pre-2004 data
158
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Peru
Organismo Supervisor de Inversión Privada en Telecomunicaciones (OSIPTEL)
Website: http://www.osiptel.gob.pe
1994 Decreto Legislativo 702
2
Legislature
Other Ministry
Other: Chair of the
Council of
Ministers,
Inspectorate
General of the
Republic, Ministry
of Economics and
Finance
turnover: 100%
Yes: 5 Members
Philippines
National Telecommunications Commission
1979 EO546
Website: http://www.ntc.gov.ph
Sector Ministry
(annual report)
Yes: 3 Members
Poland
Office of Telecommunications and Post Regulation (URTIP)
2000 Law of July 2000
Sector Ministry
(annual report)
Website: http://www.urtip.gov.pl
Other: 100% State budget in the
amount fixed each year in the
Annual Budget Act.
No: President
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
159
WORLD
Country
Name of Authority
Year
created
the authority
1
Financed by
Reports to
Portugal
National Communications Authority (ANACOM)
1989 Decree-Law No. 188/81 of 2
July. Decree-law No. 283/89
of 23 August, now revoked
by Decree-law 309/2001
of 7 December
Website: http://www.anacom.pt
Legislature
No reporting
requirements
Other: Annual
report to both
Government and
Parliament. The
chairman of the
board of
administration will
respond to
requests for
hearing addressed
by the appropriate
committee of the
Parliament, to
provide information
or clarification on
its activities.
Licence fees: 0.0014%
Spectrum fees: 97%
Yes: 3 Members
Other: 2.9% Laboratory tests,
postal services, EU subsidies,
extraordinary income
Qatar
Supreme Council for Communication and Information Technology
Website: http://www.ict.gov.qa/en/Default.aspx
2004 Ameri Decree 36, year 2004
Other: Report to an Other: Not determined yet
independent board
chair by His
Highness the Heir
Apparent of the
State of Qatar
No: Secretary General
Romania
National Regulatory Authority of Romania (ANRC)
2002 Government Emergency
Other: The
Ordinance No. 79/ 2002 on
Government of
the general regulatory
Romania
framework for
communications, approved,
with amendments and
completions, by Law No.
591/2002, with subsequent
amendments and completions
Website: http://www.anrc.ro
turnover: 79%
Other: 21% contributions from
regulated postal services
No: President
1
2
2004 data
pre-2004 data
160
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
1
Financed by
Reports to
Rwanda
Agence Rwandaise de Régulation des Services d'Utilité Publique
2001 Loi n° 39/2001 du 13/09/2001 Sector Ministry
portant Création de l'Agence (annual report)
Rwandaise de Régulation
des Services d'Utilité Publique
Website: http://www.rura.gov.rw/
Licence fees: 6,3%
Yes: 7 Members
turnover: 50,7%
Other: 43% World Bank project
and State subsidies
Samoa
Minstry of Communications and Information Technology (MCIT) (ad interim)
2005 Telecommunications Act 2005 Legislature
Other: Minister
Award/auction of mobile licence
Licence fees
Spectrum fees
Numbering fees
Regulatory fees
Fines/penalties
No: Regulator
Financial income
Saudi Arabia
Communication and Information Technology Commission
2002 The Council of ministers
decision No. (74) Dated 2001
1
Sector Ministry
(annual report)
Website: http://www.citc.gov.sa
Yes: 8 Members
Senegal
Agence de Régulation des Télécommunications
2001
Website: http://www.art.sn
Other: President of Spectrum fees: 100%
the Republic
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
161
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Singapore
Infocomm Development Authority of Singapore
1992 Telecommunication Authority
of Singapore Act 1992 and
subsequently superseded by
the Info-communications
Development Authority of
Singapore Act 1999
Website: http://www.ida.gov.sg
Other: Annual
Report to the
Minister for
Information,
Communications
and the Arts.
Other: financed from a
combination of licence fees,
numbering fees and spectrum
fees
No: CEO
Slovak Republic
Telecommunication Office
Website: http://www.teleoff.gov.sk
2000 Act. No. 195 / 2000 C.l. on
Telecommunications
Other: Annual
report to the
National Council of
the Slovak republic
No: President
Slovenia
Post and Electronic Communications Agency (APEK)
2001 Government's Decision on
establishment of
Telecommunication and
Broadcasting Agency
1
Website: http://www.apek.si
Other: Annual
Report to the
Government and
the National
Assembly of the
Republic of Slovenia
Licence fees: 6%
Spectrum fees: 37%
Numbering fees: 46%
Fines/penalties: 1%
No: M. Sc., Acting
Director
Other: 10% Postal fees
South Africa
ICASA
Website: http://www.icasa.org.za
2000 ICASA Act 13 of 2000
Legislature
Sector Ministry
(annual report)
Yes: 7 Members
1
2
2004 data
pre-2004 data
162
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Spain
Comisión del Mercado de las Telecomunicaciones (CMT)
1996 Real Decreto-Ley 6/1996, de
7 de junio, de liberalización
de las Telecomunicaciones.
Actualmente se regula en la
Ley 32/2003, General de
Telecomunicaciones
1
Legislature
Other: To the
Government, for
submission to
Parliament
Website: http://www.cmt.es
Yes: 9 Members
turnover: 0.15%
Other: Various telecommunication
fees, from the issue of
registration certificates, technical
opinions and inspection or
monitoring activities.
Sri Lanka
Telecommunications Regulatory Commission of Sri Lanka
1991 Sri Lanka Tel. Act No. 25 of
Sector Ministry
1992 & Tel. (Amendment Act) (annual report)
No. 27 of 1996
Website: http://www.trc.gov.lk
Licence fees: 30%
Spectrum fees: 51%
Yes: 5 Members
St. Lucia
National Telecommunications Regulatory Commission
2000 Telecommunications Act 2000 Sector Ministry
(annual report)
Website: http://www.ntrc.org.lc
Spectrum fees: 100%
Yes: 5 Members
St. Vincent and the Grenadines
National Telecommunications Regulatory Commission (NTRC)
2001 Telecommunications Act No.1 Sector Ministry
2001
(annual report)
Website: http://www.ntrc.vc
Yes: 5 Members
Other: 2.0% Application fees and
rental of conference room
Sudan
National Telecommunication Corporation (NTC)
1994 Telecommunication Act 2001
(including establishment of
Regulatory Authority)
Website: http://www.ntc.org.sd
Sector Ministry
(annual report)
Licence fees: 33.3%
Spectrum fees: 50%
Yes: 10 Members
turnover: 16.7%
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
163
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
Suriname
Telecommunications Authority Suriname (TAS)
1998 Decree 1998 Establishment
of the Telecommunications
Authority Suriname
Website: http://www.tas.sr
Sector Ministry
(annual report)
Other: Board of
Directors
No: Managing Director
Sweden
National Post and Telecom Agency
1992 Förordning med instruktion
för telestyrelsen (Decree)
Website: http://www.pts.se/
Sector Ministry
(annual report)
Licence fees: 41%
Other: 18%
Switzerland
Office Fédéral de la Communication (OFCOM)
1992 Loi du 30 avril 1997 sur les
télécommunications (LTC)
1
Website: http://www.ofcom.ch
Legislature
Sector Ministry
(annual report)
No: Director
Tanzania
Tanzania Communications Regulatory Authority
Website: http://www.tcc.go.tz
Sector Ministry
1994 Tanzania Communications
Act No. 18 of 1993, Tanzania (annual report)
Communications Regulatory
Authority Act No. 12 of 2003
(merger of TCC and TBC in
2003)
Licence fees: 2.37%
Yes: 7 Members
turnover: 52.39%
Other: 2.08% Applications forms
and Type approval fee
1
2
2004 data
pre-2004 data
164
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Financed by
Reports to
TFYR Macedonia
Agency for electronic communications
2005 Law on electronic
communications
1
Website: http://www.aec.mk
Other: Annual
report to the
Parliament
Licence fees: 8,5%
Numbering fees: 11,6%
Yes: 5 Members
Thailand
National Telecommunications Commission (NTC)
2004 The Act on Organizations to
Assign Radio Frequency
Sprectrum and to Regulate
the Sound Broadcasting,
Television Broadcasting and
Telecommunication Services,
B.E.2543
Website: http://www.ntc.or.th
Other: Report to the
Cabinet
Yes: 7 Members
Togo
Autorité de Réglementation des Secteurs de Postes et Télécommunications (ART&P)
Website: http://www.artp.tg
1998 Loi n° 98-005 du 11 février
1998 en son article 57
Sector Ministry
(annual report)
Yes: 5 Members
Trinidad and Tobago
Telecommunications Authority of Trinidad and Tobago
2004 Telecommunications Act
2001 (as amended by
Telecommunications
(Amendment) Act 2004
Sector Ministry
(annual report)
Website: http://www.tatt.org.tt
Licence fees: 40%
Yes: 11 Members
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
165
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Tunisia
Instance Nationale des Télécommunications
Website: http://www.intt.tn/
2001 Loi n° 2001-1 du 15 janvier
Legislature
2001, portant promulgation du Sector Ministry
code des télécommunications (annual report)
Numbering fees: 100%
Yes: 7 Members
Turkey
Telecommunications Authority
Website: http://www.tk.gov.tr
2000 Amending Law No. 4502
Other: Turkish
National Assembly
Spectrum fees: 88%
Yes: 7 Members
turnover: 9%
Other: 3% caution returns; form,
book, publication, hologram sales
returns, etc.
Uganda
Uganda Communications Commission
Website: http://www.ucc.co.ug
1997 The Communications Act 1997 Legislature
Sector Ministry
(annual report)
Licence fees: 4%
Spectrum fees: 47%
Yes: 7 Members
turnover: 33%
Other: 14% Rent
United Arab Emirates
Telecommunications Regulatory Authority
2004 Federal Law by decree
No. (3) of 2003
Website: http://www.tra.ae
Other: UAE
Telecom Supreme
Committee
Other: 100% UAE Federal
Governmet
Yes: 5 Members
1
2
2004 data
pre-2004 data
166
R EGULATORY T ABLES
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
United Kingdom
Office of Communications (OFCOM)
1984 Telecommunications Act
1984; 2003 Communications
Act
1
Licence fees: 16%
No: Chief Executive
turnover: 15%
Other: 5% Government
department (DCMS and DTI),
commercial property sub-lets and
broadcasting application fees and
commercial interference income
United States
Federal Communications Commission
1934 The Communications Act of
1934
2
Website: http://www.ofcom.org.uk
Legislature
Sector Ministry
(annual report)
Other: Select
Committees of
Legislature (Public
Accounts
Committee, Trade &
Industry Select
Committee). The
annual reports to
the sector Ministry
is also required to
be provided to the
Legislature by the
Minister.
Website: http://www.fcc.gov/
Legislature
Other: It is
independent with
Congressional
oversight and
budget control
Yes: 5 Members
Other: Auctions receipts are
deposited with Treasury. Funds
are allocated from receipts as
needed to cover the cost of
running the Auctions Program.
Uruguay
Unidad Reguladora de Servicios de Comunicaciones (URSEC)
2001 Ley N° 17.296 del 21 de
febrero de 2001
Website: http://www.ursec.gub.uy
Other: The reports Spectrum fees: 78%
are sent to
whoever asks for Fines/penalties: 1%
them. The
Executive Authority
must receive
reports on matters
falling within its
competence.
Yes: 3 Members
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
167
WORLD
Country
Name of Authority
Year
created
the authority
Reports to
Financed by
Venezuela
Comisión Nacional de Telecomunicaciones (CONATEL)
1991 Decreto n° 1826 del 5 de
septiembre de 1991
Sector Ministry
(annual report)
Website: http://www.conatel.gov.ve
Yes: 8 Members
Other: 82.2% Special contribution
Zambia
Communications Authority
Website: http://www.caz.gov.zm
1994 Telecommunications Act of
1994 Cap 469
Legislature
Sector Ministry
(annual report)
Licence fees: 6%
Yes: 8 Members
turnover: 77.4%
Other: 0.8% Rentals
Zimbabwe
Postal and Telecommunications Regulatory Authority of Zimbabwe (POTRAZ)
Website: http://www.potraz.gov.zw/
2000 Postal and
Telecommunications Act
2000, Chapter 12:05
Legislature
Sector Ministry
(annual report)
Licence fees: 17%
Spectrum fees: 5%
Yes: 7 Members
turnover: 47%
Other: 3% Postal
1
2
2004 data
pre-2004 data
168
R EGULATORY T ABLES
2. Status of the main fixed-line operators
Name of the operator
WORLD
Status
Afghanistan
Afghan Telecom
State-owned - Corporatized
The government intends to
privatize the operator
(Year: 2006)
Albania
Albtelecom sh.a
In the process of privatizing
Algeria
Algérie Telecom
(Year: 2006)
2
Andorra
Servei de Telecomunicacions
d’Andorra (STA)
1
Angola
Angola Telecom
State-owned
Mercury
Partially privatized
Mundo Startel
Fully privatized
Nexus
Fully privatized
Wezacom
Fully privatized
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
169
Antigua and Barbuda
2
Argentina
Armenia
Australia
1
Austria
WORLD
Status
Antigua Public Utilities Authority
Telephones
State-owned
Telecom
Fully privatized
Date of privatization (first
phase): 1990
Telefónica
Fully privatized
phase): 1990
ArmenTel JV
Partially privatized: 90%
phase): 1998: 90% sold
Telstra
Partially privatized: 51.8%
phase): 1997: 33.3% sold
Telekom Austria
1
2
2004 data
pre-2004 data
170
R EGULATORY T ABLES
Azerbaijan
Bahamas
Bahrain
Bangladesh
Barbados
WORLD
Status
AzTelekom
State-owned
(Year: 2008)
Baktelekom
State-owned
(Year: 2008)
Bahamas Telecommunications
Company Ltd
Batelco
Bangladesh Telegraph and
Telephone Board (DTTB)
State-owned
Cable & Wireless (Barbados) Ltd
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
171
Belarus
2
2
2
2
WORLD
Status
Republican Unitary Enterprise
No intention to privatize at
present
Belgacom
Partially privatized: 49,9%
phase): 1996: 49,9% sold
Belize Telecommunications Ltd
Benin
Office des postes et
télécommunications (OPT)
State-owned
(Year: 2002/2003)
Bhutan
Bhutan Telecom
present
ENTEL S.A.
Fully privatized
phase): 1995: 100% sold
Belgium
Belize
Bolivia
1
2
2004 data
pre-2004 data
172
R EGULATORY T ABLES
2
Bosnia and
Herzegovina
Botswana
2
1
Brazil
Brunei Darussalam
Bulgaria
WORLD
Status
BH Telecom
(Year: 2004)
HT Mostar
Telekom Srpske
Botswana Telecommunications
Corporation
Empresa Brasileira de
Telecomunicações S.A. (Embratel)
Fully privatized
phase): 1998
Brasil Telecom
Fully privatized
phase): 1998
CTBC - Companhia Telefônica Brasil
Central
Fully privatized
phase): 1998
Telemar
Fully privatized
phase): 1998
JTB
State-owned
Bulgarian Telecomunication
Company BTC
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
173
WORLD
Status
1
Burkina Faso
ONATEL
2
Burundi
Office National des
Télécommunications
2
Cambodia
Ministry of Posts and
Telecommunications
State-owned
1
Cameroon
CAMTEL
Canada
SaskTel
State-owned
present
Bell Canada
Fully privatized
TELUS
Fully privatized
phase): 1990: 100% sold
MTS
Fully privatized
phase): 1994: 100% sold
Aliant
Fully privatized
1
2
2004 data
pre-2004 data
174
R EGULATORY T ABLES
Cape Verde
2
1
Central African Rep.
Chad
Chile
China
WORLD
Status
CVTelecom
phase): 1995: 53.7% sold
Socatel
SOTEL TCHAD
CTC
Fully privatized
VTR
Fully privatized
EntelPhone
Fully privatized
Telefónica del Sur
Fully privatized
China Telecom
phase): 2002
China Unicom
phase): 2000
China Netcom
phase): 2004
China Mobile
phase): 1997
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
175
Colombia
WORLD
Status
Colombia Telecomunicaciones
ETB
EPM
UNITEL
Fully privatized
ETELL
present
2
Comoros
SNPT
present
2
Congo
ONPT
(Year: 1965)
1
Congo
(Dem. Rep. of)
OCPT
RENATELSAT
Instituto Costarricense de
Electricidad
State-owned
present
Costa Rica
1
2
2004 data
pre-2004 data
176
R EGULATORY T ABLES
1
1
Côte d'Ivoire
Croatia
Cuba
Cyprus
Czech Republic
2
D. P. R. Korea
WORLD
Status
Côte d'Ivoire Télécom
Croatian Telecom Inc. (HT)
ETECSA
Fully privatized
Cyprus Telecommunications
Authority (CYTA)
present
CESKÝ TELECOM, a.s.
Fully privatized
Ministry of Posts and
Telecommunications
State-owned
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
177
Denmark
2
Djibouti
2
Dominica
1
Dominican Rep.
Ecuador
Egypt
WORLD
Status
TDC
Fully privatized
OPT
(Year: 1957)
Telecommunications of Dominica
VERIZON Dominicana
Fully privatized
TRICOM, S.A.
Fully privatized
ALL America Cables & Radio, Inc.
(AAC&R- Centennial Dominicana)
Fully privatized
Andinatel S.A.
present
Pacifictel S.A.
present
Telecom Egypt
1
2
2004 data
pre-2004 data
178
R EGULATORY T ABLES
El Salvador
2
Equatorial Guinea
Eritrea
Estonia
1
Ethiopia
2
Fiji
WORLD
Status
CTE
TELEFONICA
TELEMOVIL
Fully privatized
GCA
Fully privatized
Sociedad Anónima de
Telecomunicaciones de la República
de Guinea Ecuatorial (Getesa)
ERITEL
Elion Enterprises
Ethiopian Telecommunication
Corporation
Telecom Fiji Limited
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
179
1
Finland
WORLD
Status
TeliaSonera AB
ElisaCom Oy
Fully privatized
France Télécom
Gabon
GABON TELECOM
State-owned
Gambia
GAMTEL
Georgia
GEC
New Nets
Fully privatized
01051 Telecom GmbH
Fully privatized
Arcor AG & Co.
Partially privatized: ~82%
Colt Telecom GmbH
Fully privatized
Deutsche Telekom AG
Partially privatized: ~62.5%
France
Germany
phase): 1998
(Year: 2004)
phase): 1996: ~26% sold
1
2
2004 data
pre-2004 data
180
R EGULATORY T ABLES
1
Ghana
Greece
2
2
2
1
Grenada
Guatemala
Guinea
Guinea-Bissau
WORLD
Status
Ghana Telecom
WESTEL
Fully privatized
OTE (Hellenic Telecommunications
Organisation S.A )
Tellas
Forthnet
Grenada Telecommunications
(GRENTEL)
Empresa Telecomunicaciones de
Guatemala S.A.
Fully privatized
phase): 1997: 100% sold
Sotelgui
Guiné Telecom
phase): 1997: 7.5% sold
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
181
2
Guyana
Haiti
2
Honduras
Hungary
2
Iceland
India
WORLD
Status
Guyana Telephone and Telegraph
Ltd (GT & T)
phase): 1991
Téléco
Haitel
Fully privatized
Empresa Hondureña de
Telecomunicaciones (HONDUTEL)
T-Com
Fully privatized
phase): 1993: 33.2% sold
Invitel
Fully privatized
Hungarotel
Fully privatized
Emitel
Fully privatized
Iceland Telecom Ltd
BSNL
MTNL
Reliance
Fully privatized
Bharti (Airtel)
Fully privatized
1
2
2004 data
pre-2004 data
182
R EGULATORY T ABLES
1
2
Indonesia
2
Status
PT. Telkom
PT. Indosat
Iran (I.R.)
TCI
Iraq
Iraqi Telecommunications and Posts
State-owned
Eircom
Fully privatized
BT Ireland
Fully privatized
Bezeq
Telecom Italia
Ireland
2
WORLD
Israel
Italy
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
183
Jamaica
Japan
Jordan
2
Kazakhstan
1
Kenya
2
Kiribati
WORLD
Status
Cable and Wireless
NTT EAST
Fully privatized
phase): 1999: 100% sold
NTT WEST
Fully privatized
phase): 1999: 100% sold
Jordan Telecom (JT)
Kazakhtelecom
phase): 1994
Telkom Kenya Limited
(Year: 2006)
Telecom Services Kiribati Limited
1
2
2004 data
pre-2004 data
184
R EGULATORY T ABLES
Korea (Rep.)
Kuwait
Kyrgyzstan
Lao P.D.R.
Latvia
WORLD
Status
Korea Telecom
Fully privatized
Ministry of Communications (MOC)
State-owned
Kyrgyztelecom
Saima Telecom
Fully privatized
Winline
Fully privatized
Instrumentaalshik
Fully privatized
Enterprise of Telecom Lao
State-owned
Lao Telecom Co. Ltd
Latvian Railway (LDz)
State-owned
present
Latvenergo
State-owned
present
Lattelekom
present
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
185
2
Lebanon
1
Lesotho
WORLD
Status
Ministry of Telecommunications
State-owned
Telecom Lesotho
phase): 2000
2
Liberia
Liberia Telecommunications
Corporation
2
Libya
General Post and
State-owned
Telecommunication Company (GPTC)
2
Liechtenstein
LTN Lie. TeleNet
State-owned
present
Telecom FL
State-owned
present
AB Lietuvos Telekomas
Lithuania
1
2
2004 data
pre-2004 data
186
R EGULATORY T ABLES
Luxembourg
Enterprises of the P&T (EPT)
WORLD
Status
present
TELECOM MALAGASY
Malawi Telecommunications Ltd
Telekom Malaysia
phase): 1990
Dhiraagu
phase): 1988
SOTELMA
(Year: 2006)
IKATEL
Fully privatized
See www.ilr.lu
Madagascar
2
Malawi
Malaysia
2
Maldives
Mali
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
187
Malta
2
1
Marshall Islands
Mauritania
Mauritius
Mexico
2
Micronesia
WORLD
Status
Maltacom plc
National Telecommunications
Authority
Mauritel S.A
Mauritius Telecom Ltd
Teléfonos de México
Fully privatized
phase): 1991: 100% sold
FSM Telecommunications
Corporation
(Year: 1981)
1
2
2004 data
pre-2004 data
188
R EGULATORY T ABLES
Moldova
2
Monaco
Mongolia
Morocco
Mozambique
Myanmar
WORLD
Status
Moldtelecom S.A.
present
MONACO TELECOM
Fully privatized
Mongolian Telecommunications
Company
Ittisalat Al-Maghrib
TDM
Myanma Posts and
Telecommunications
State-owned
Bagan Cybertech
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
189
2
WORLD
Status
Namibia
Telecom Namibia
State-owned
Nauru
Directorate of Telecommunications
State-owned
Nepal
Nepal Telecom
State-owned
(Year: After 2006)
KPN
Telecom Corporation of New
Zealand
Fully privatized
phase): 1990: 100% sold
ENITEL
Fully privatized
Netherlands
New Zealand
Nicaragua
present
1
2
2004 data
pre-2004 data
190
R EGULATORY T ABLES
2
2
Niger
Nigeria
Norway
Oman
Pakistan
1
Panama
WORLD
Status
SONITEL
NITEL
Telenor
Omantel
Pakistan Telecom Company Limited
(PTCL)
Cable & Wireless Panama
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
191
Status
2
Papua New Guinea
Telikom PNG Limited
State-owned
1
Paraguay
COPACO
Telefónica del Perú S.A.
Fully privatized
Telmex
Fully privatized
BellSouth Perú S.A.
Fully privatized
Americatel Perú S.A.
Fully privatized
All fixed-line operators
Fully privatized
Telekomunikacja Poska
PT Comunicações, S.A.
Peru
2
Philippines
Poland
1
Portugal
WORLD
(Year: 2003)
phase): May-1994: 35% sold
phase): 1995: 27.26% sold
1
2
2004 data
pre-2004 data
192
R EGULATORY T ABLES
Qatar
1
2
Status
Q-Tel
S.C. Romtelecom S.A.
Svyazinvest is the government
controlled telecommunications
holding company (an umbrella
organization).
Rwandatel
Principe
Companhia Santomense de
Telecomunicações, SARL
Samoa
SamoaTel
Romania
2
WORLD
Russia
Rwanda
S. Tomé and
(Year: 2006)
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
193
2
San Marino
Saudi Arabia
1
2
Senegal
Serbia and
Montenegro
Seychelles
2
Sierra Leone
WORLD
Status
Telecom Italia
Saudi Telecom
SONATEL
phase): 1996: 63,33% sold
Telecom Serbia
phase): 1997
Cable and Wireless (Seychelles) Ltd
Fully privatized
Telecom (Seychelles) Ltd
Fully privatized
Sierratel
(Year: 2004)
1
2
2004 data
pre-2004 data
194
R EGULATORY T ABLES
Singapore
Slovak Republic
Slovenia
2
2
1
Solomon Islands
Somalia
South Africa
WORLD
Status
Singapore Telecommunications Ltd
phase): 1993: 22.2% sold
Slovak Telecom a.s.
Telekom Slovenije
phase): 1996
Solomon Telekom Company Ltd.
Somtel
Fully privatized
Telkom
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
195
Spain
1
Sri Lanka
WORLD
Status
Telefónica
Fully privatized
phase): 1997
Auna
Fully privatized
phase): 1997
Grupo Ono
Fully privatized
Euskaltel
Sri Lanka Telecom Ltd
Suntel Ltd
Fully privatized
Lanka Bell (Pvt) Ltd
Fully privatized
Cable & Wireless (West Indies)
Limited
Fully privatized
Cable and Wireless (WI) Limited
Fully privatized
SUDATEL
KANARTEL
Fully privatized
phase): 2005: 100% sold
St. Lucia
St. Vincent and the
Grenadines
Sudan
1
2
2004 data
pre-2004 data
196
R EGULATORY T ABLES
2
Status
Suriname
Telesur
Swaziland
Swaziland Posts &
Telecommunications Corporation
TeliaSonera AB
phase): 2000: 29.4% sold
Swisscom AG
S.T.E.
State-owned
present
Tajiktelecom
Sweden
Switzerland
Syria
2
WORLD
Tajikistan
present
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
197
1
Tanzania
TFYR Macedonia
1
Thailand
WORLD
Status
Tanzania Telecommunications
Company Ltd (TTCL)
A.D. Makedonski telekomunikacii
TOT Corporation Plc., Ltd
TRUE Corporation Plc., Ltd
TT&T Plc. Ltd
2
Togo
Togo Télécom
Tonga
Tonga Telecommunications
Corporation Limited (TCC)
Cable & Wireless
Trinidad and Tobago
1
2
2004 data
pre-2004 data
198
R EGULATORY T ABLES
WORLD
Status
Tunisia
Tunisie Télécom
Turkey
Turk Telekom
2
Turkmenistan
Turkmentelecom
(Year: 1992)
2
Tuvalu
Tuvalu Telecom Corporation
(Year: 1994)
present
Uganda Telecom
MTN
Fully privatized
phase): 1998: 100% sold
Ukrainian Telecom Corporation
Uganda
2
Ukraine
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
199
United Arab Emirates
United Kingdom
1
United States
WORLD
Status
Etisalat
British Telecommunications Plc
Fully privatized
Kingston Plc
Fully privatized
All fixed-line operators
Fully privatized
2
Uruguay
ANTEL
State-owned
2
Uzbekistan
Halkapo and Machalit
2
Vanuatu
Telecom Vanuatu Ltd
present
1
2
2004 data
pre-2004 data
200
R EGULATORY T ABLES
Venezuela
Status
CANTV
MOVISTAR
Fully privatized
DIGITEL
Fully privatized
INFONET
Fully privatized
VNPT
State-owned
VIETTEL, ETC, VISHIPEL
State-owned
SPT
HANOI TELECOM
Yemen
Public Telecom Corporation
present
Zambia
Zamtel
State-owned
present
Zimbabwe
Tel One
TELEACCESS
Fully privatized
Viet Nam
1
WORLD
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
201
3. Level of competition
Local
Long Int'l WLL
services distance
WORLD
Data
DSL
Cable VSAT Leased FWB Mo- Pagmodem
lines
bile ing
Cable
TV
FSS MSS GMPCS IMT- IS IG
2000
Afghanistan
C
P
P
P
...
...
...
P
P
...
P
...
P
P
...
P
...
P
P
Albania
C
P
P
P
C
C
...
C
C
P
P
C
C
C
...
C
...
C
P
Algeria
P
P
P
P
C
C
...
P
C
P
P
...
...
...
...
P
...
C
P
2
Andorra
M
M
M
...
M
...
...
...
M
...
M
M
D
...
M
...
...
...
...
1
Angola
C
C
C
C
C
...
C
...
C
C
P
C
P
...
C
C
...
C
C
Antigua and
Barbuda
M
...
M
C
C
M
C
C
M
C
C
C
C
...
...
...
...
C
M
Argentina
C
C
C
C
C
C
...
C
C
...
C
C
C
C
C
C
...
C
...
Armenia
M
M
M
...
C
C
...
...
M
C
P
M
C
C
C
C
...
C
...
Australia
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Austria
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Azerbaijan
P
M
P
P
C
C
C
C
P
C
P
P
C
C
...
P
...
C
M
Bahamas
P
P
P
...
P
C
M
C
P
...
M
C
M
...
...
...
...
C
M
Bahrain
C
...
C
...
C
C
...
C
C
...
P
C
...
...
...
...
P
C
C
Bangladesh
C
C
M
P
C
C
...
C
...
...
C
M
...
M
...
...
...
C
M
Barbados
M
...
P
P
...
...
...
C
M
...
C
C
C
...
...
...
...
C
P
Belarus
C
M
M
...
C
C
C
...
M
...
P
C
C
M
M
M
...
C
M
2
1
Key: WLL = Wireless local loop; DSL = Digital subscriber line; VSAT = Very small aperture terminal; FWB = Fixed wireless broadband;
FSS = Fixed-satellite service; MSS = Mobile-satellite service; GMPCS = Global Mobile Personal Communications System; IS = Internet services;
IG = International gateways;
M = Monopoly; D = Duopoly; P = Partial competition; C = Full competition; ... = Not available
1
2
2004 data
pre-2004 data
Note: This table reflects what is legally permissible; therefore it may not reflect the actual number of operators in the market.
R EGULATORY T ABLES
203
Local
Long Int'l WLL
services distance
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
2
Belgium
C
C
C
P
C
C
C
C
C
...
P
P
...
C
C
C
P
C
...
2
Belize
M
M
M
M
M
...
...
...
M
...
...
M
C
...
...
P
...
M
...
2
Benin
M
M
M
...
...
M
...
...
M
...
C
C
...
...
...
...
...
...
...
Bhutan
M
M
M
M
P
...
...
P
M
M
M
M
C
...
M
M
...
P
M
2
Bolivia
M
M
M
...
C
...
...
...
M
...
C
...
C
M
M
...
...
C
...
2
Bosnia and
Herzegovina
M
M
M
...
...
...
...
...
...
...
P
...
C
...
...
...
...
C
...
Botswana
M
C
M
M
C
C
...
C
C
C
C
C
...
...
...
...
...
C
M
2
Brazil
C
C
C
C
C
C
C
C
C
...
C
C
C
C
C
C
C
C
...
1
Brunei
Darussalam
P
P
P
...
P
...
...
...
P
...
M
M
...
...
...
...
...
P
P
Bulgaria
P
P
P
C
C
C
...
C
C
P
P
C
C
C
...
...
P
C
...
1
Burkina Faso
M
M
M
C
C
C
C
P
P
C
C
C
C
P
P
P
P
C
M
2
Burundi
C
C
C
C
C
C
...
C
C
...
C
C
C
...
C
C
C
C
...
2
Cambodia
P
P
P
P
P
...
...
P
P
...
P
P
P
...
...
...
...
P
...
1
Cameroon
M
M
M
C
C
...
...
P
...
...
C
C
C
...
...
C
...
C
...
Canada
C
C
C
C
C
C
C
C
C
C
C
C
...
C
C
C
C
C
C
Cape Verde
M
M
M
M
C
C
C
P
...
C
C
C
C
C
P
C
C
C
M
1
2
2004 data
pre-2004 data
204
R EGULATORY T ABLES
Local
Long Int'l WLL
services distance
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
2
Central
African Rep.
M
M
M
...
C
...
...
...
C
...
C
...
C
...
...
C
...
...
...
1
Chad
M
C
M
...
C
...
...
C
...
...
...
...
...
C
C
...
...
C
...
Chile
P
C
C
M
C
P
M
C
P
M
C
...
M
C
M
M
...
C
C
China
P
P
P
P
P
P
...
C
P
...
P
C
...
M
M
...
...
C
...
Colombia
P
C
C
...
C
...
...
...
...
...
P
C
C
...
...
...
...
C
...
2
Comoros
M
M
M
M
M
...
...
...
M
...
...
...
...
M
...
...
...
M
...
2
Congo
C
C
C
...
...
...
...
...
...
...
C
C
...
...
C
...
...
...
...
1
Congo (Dem.
Rep. of)
C
C
P
P
C
C
P
C
P
...
C
P
C
C
C
C
C
C
...
Costa Rica
M
M
M
M
M
M
P
M
M
M
M
C
P
M
M
M
M
M
...
1
Côte d'Ivoire
P
P
P
C
C
M
...
C
P
P
P
C
...
C
...
C
...
C
...
1
Croatia
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Cuba
M
M
M
M
M
M
M
M
M
...
M
M
P
M
M
M
...
P
...
Cyprus
C
C
C
...
C
M
...
...
C
...
P
M
P
C
C
...
...
C
...
Czech
Republic
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
D.P.R. Korea
M
M
M
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
2
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
205
Local
Long Int'l WLL
services distance
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
Denmark
C
C
C
C
C
C
C
C
C
P
P
...
C
C
C
C
P
C
C
2
Djibouti
M
M
M
...
M
...
...
...
M
...
...
M
...
M
M
...
...
M
...
2
Dominica
M
M
M
...
M
...
...
...
M
...
M
M
C
M
M
...
...
...
...
1
Dominican
Rep.
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Ecuador
P
P
P
P
C
C
C
C
C
C
P
C
C
C
C
...
P
C
...
Egypt
M
M
M
M
C
C
...
C
M
...
P
...
...
...
...
C
...
C
M
El Salvador
C
C
C
C
C
C
C
C
C
...
C
C
C
C
C
C
...
C
...
Equatorial
Guinea
M
M
M
...
M
...
...
...
...
...
...
...
C
M
M
...
...
M
...
Eritrea
M
M
M
M
P
...
...
P
M
M
P
...
M
M
P
...
...
P
M
Estonia
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
1
Ethiopia
M
M
M
M
M
...
...
M
M
M
M
...
...
M
M
M
...
M
M
2
Fiji
M
M
M
...
M
...
...
...
M
...
M
C
M
M
C
...
...
...
...
1
Finland
C
C
C
C
C
C
C
C
C
...
P
...
C
C
C
C
P
C
C
France
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Gabon
M
C
C
C
C
C
M
C
C
...
C
...
...
M
C
C
M
C
...
Gambia
M
M
M
M
C
M
...
M
M
...
P
...
...
...
...
...
...
C
M
2
1
2
2004 data
pre-2004 data
206
R EGULATORY T ABLES
Local
Long Int'l WLL
services distance
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
Georgia
P
P
C
P
C
P
C
C
M
...
C
...
C
P
P
C
...
C
...
Germany
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Ghana
P
P
P
P
C
P
C
P
P
...
P
C
C
C
C
C
P
C
...
Greece
C
C
C
C
C
C
C
C
C
...
P
P
C
C
P
P
P
C
C
2
Grenada
M
M
M
...
M
...
...
...
M
...
M
M
...
M
M
...
...
...
...
2
Guatemala
C
M
C
C
C
...
...
C
C
...
C
C
C
C
C
...
...
C
...
2
Guinea
P
P
P
...
...
...
...
...
...
...
P
C
C
C
C
C
...
C
...
1
Guinea-Bissau M
M
M
C
C
C
C
C
M
C
P
C
C
C
C
C
P
C
P
2
Guyana
M
M
M
...
D
...
...
...
M
...
C
C
...
...
...
...
...
...
...
Haiti
P
P
P
...
C
...
...
C
...
...
P
...
C
...
...
...
...
C
...
Honduras
M
M
M
...
C
...
...
C
P
...
M
C
C
C
C
C
...
...
...
Hungary
C
C
C
P
C
...
C
C
C
C
P
...
C
C
C
C
P
...
C
Iceland
C
C
C
C
C
C
...
C
C
...
C
M
...
...
...
...
...
C
...
India
C
C
C
C
C
...
C
C
...
...
C
C
C
...
...
...
...
C
C
Indonesia
P
P
P
P
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Iran (I.R.)
P
M
M
C
P
C
...
M
M
P
P
C
...
M
M
M
...
P
M
Iraq
M
M
M
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
1
2
2
1
2
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
207
Local
Long Int'l WLL
services distance
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
Ireland
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
2
Israel
M
M
C
C
...
...
...
C
M
...
C
C
M
C
C
...
...
...
...
2
Italy
C
C
C
C
C
C
...
C
C
...
C
C
C
C
C
C
C
C
...
Jamaica
C
C
C
C
C
C
C
C
C
C
C
...
P
...
...
...
...
C
C
Japan
C
C
C
C
...
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Jordan
P
P
P
C
C
C
C
C
C
C
P
C
C
C
...
C
...
C
C
2
Kazakhstan
C
C
C
...
C
...
...
...
C
...
D
C
C
C
...
...
...
...
...
1
Kenya
P
P
P
C
C
C
...
P
...
C
P
C
...
...
...
P
...
C
P
2
Kiribati
M
M
M
...
M
...
...
...
M
...
...
...
...
...
...
...
...
...
...
Korea (Rep.)
C
C
C
C
C
C
C
...
C
...
C
C
...
...
...
P
C
C
...
Kuwait
M
...
M
...
P
...
...
P
M
M
...
P
P
M
...
M
...
P
M
Kyrgyzstan
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
...
C
C
Lao P.D.R.
P
P
P
P
P
P
...
...
...
P
P
...
...
...
...
...
P
P
M
Latvia
C
C
C
...
C
C
C
C
C
...
C
C
C
...
...
...
...
C
...
2
Lebanon
M
M
M
...
C
...
...
...
...
...
...
...
...
M
...
...
...
...
...
1
Lesotho
P
P
P
P
P
...
...
P
M
...
C
...
...
...
...
...
...
C
M
2
Liberia
P
P
C
M
P
...
...
P
...
...
C
...
...
...
...
...
...
...
...
1
2
2004 data
pre-2004 data
208
R EGULATORY T ABLES
Local
Long Int'l WLL
services distance
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
2
Libya
M
M
M
...
...
...
...
...
M
...
M
...
...
M
M
...
...
...
...
2
Liechtenstein
C
...
C
...
C
C
P
...
C
...
P
M
C
...
...
...
P
C
...
Lithuania
C
C
C
C
C
C
C
C
C
C
C
...
C
C
C
C
...
C
C
Luxembourg
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Madagascar
M
M
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Malawi
M
M
P
...
...
...
...
P
...
...
P
...
...
...
...
...
...
P
...
Malaysia
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Maldives
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
Mali
P
P
P
C
P
...
P
P
P
C
P
...
...
P
P
...
...
C
...
Malta
C
...
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
2
Marshall
Islands
M
M
M
...
M
...
...
...
M
...
...
...
C
M
M
...
...
...
...
1
Mauritania
M
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Mauritius
C
...
C
C
C
C
...
C
C
C
C
C
...
C
C
C
C
C
C
Mexico
C
C
C
C
C
P
C
C
C
C
C
C
C
C
C
C
...
C
...
Micronesia
M
M
M
...
M
...
...
...
M
...
C
C
C
C
C
...
...
...
...
Moldova
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
...
C
C
2
1
2
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
209
Local
Long Int'l WLL
services distance
2
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
Monaco
M
M
M
M
M
M
...
M
M
...
M
C
M
M
M
...
...
C
...
Mongolia
P
P
C
C
C
M
...
C
C
...
P
...
C
C
C
...
...
C
C
Morocco
M
M
M
M
C
M
...
C
C
...
C
C
...
C
C
C
...
C
P
Mozambique
M
M
M
...
C
M
...
C
C
M
C
C
C
...
...
C
...
C
C
Myanmar
M
M
M
P
P
P
...
P
P
P
M
...
...
M
...
...
...
P
M
Namibia
M
M
M
M
M
M
M
C
...
M
M
C
C
M
C
C
...
C
M
Nauru
M
M
M
...
M
...
...
...
M
...
...
...
...
M
M
...
...
...
...
Nepal
P
P
P
P
C
C
C
C
C
C
P
C
C
...
...
C
...
C
P
Netherlands
C
C
C
P
P
C
P
P
C
...
P
...
P
...
...
...
...
P
...
New Zealand
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Nicaragua
C
C
C
...
C
C
C
C
C
C
C
C
C
...
...
C
...
C
...
2
Niger
M
M
M
...
M
...
...
P
M
...
C
...
P
...
...
C
...
M
...
2
Nigeria
C
P
P
C
...
...
...
C
P
...
P
C
...
...
C
C
...
C
...
Norway
C
C
C
C
C
C
C
C
C
P
P
...
C
C
C
...
P
C
...
Oman
M
M
M
M
M
M
...
M
M
M
P
M
...
M
M
M
...
M
M
Pakistan
C
C
C
C
C
C
...
C
C
C
P
C
...
C
C
C
...
C
C
Panama
C
C
C
M
C
P
C
P
M
P
P
P
P
P
...
...
...
C
...
2
1
1
2
2004 data
pre-2004 data
210
R EGULATORY T ABLES
Local
Long Int'l WLL
services distance
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
2
Papua New
Guinea
M
M
M
M
M
...
...
M
M
...
M
M
M
C
M
M
M
P
...
1
Paraguay
M
M
M
C
C
...
C
C
...
...
C
C
C
C
C
C
C
C
...
Peru
C
C
C
...
C
...
...
...
...
...
C
C
C
...
C
...
...
...
...
Philippines
C
C
C
C
C
C
...
C
C
...
C
C
C
C
C
C
C
C
...
Poland
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Portugal
C
C
C
C
C
C
C
C
C
C
C
...
C
C
C
C
C
C
C
Qatar
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
Romania
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
2
Russia
D
D
D
...
C
...
...
...
M
...
C
C
C
D
D
...
...
...
...
1
Rwanda
C
C
...
...
C
...
...
C
C
...
C
...
...
...
...
C
...
C
...
2
S. Tomé and
Principe
M
...
M
M
M
...
...
M
M
...
...
...
...
M
...
...
...
...
...
Samoa
M
M
M
P
P
P
C
M
M
M
M
C
C
M
M
M
M
C
M
San Marino
M
M
M
...
D
...
...
...
D
...
C
D
...
D
...
...
...
...
...
Saudi Arabia
M
P
P
P
P
P
...
P
P
M
P
M
...
M
P
C
P
C
P
Senegal
C
C
C
C
C
C
C
C
C
C
C
...
...
...
...
...
...
C
C
2
1
2
1
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
211
Local
Long Int'l WLL
services distance
2
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
Serbia and
Montenegro
C
C
...
...
C
...
...
...
C
...
C
C
C
M
...
...
...
...
...
Seychelles
P
P
P
...
P
...
...
...
P
P
P
...
M
...
...
...
...
P
...
Sierra Leone
M
M
P
...
P
...
...
P
P
...
C
P
...
M
...
...
...
P
...
Singapore
C
...
C
C
C
C
C
C
C
C
C
C
M
C
C
C
C
C
C
Slovak
Republic
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Slovenia
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
2
Solomon
Islands
M
M
M
...
M
...
...
...
...
...
...
...
...
...
...
...
...
...
...
2
Somalia
C
C
C
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
1
South Africa
C
C
C
M
C
M
M
P
M
P
P
C
...
P
P
...
...
C
P
Spain
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Sri Lanka
P
C
P
P
C
M
...
C
P
...
C
C
...
...
...
...
...
C
...
St. Lucia
...
...
...
C
...
...
C
C
...
C
C
...
C
C
C
C
C
...
C
St. Vincent
and the
Grenadines
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Sudan
P
P
P
P
P
P
...
P
P
P
P
P
P
P
P
M
P
P
M
Suriname
M
M
M
M
M
M
M
M
M
M
M
M
P
M
M
M
M
P
M
2
1
1
2
2004 data
pre-2004 data
212
R EGULATORY T ABLES
Local
Long Int'l WLL
services distance
2
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
Swaziland
M
M
M
M
M
...
...
C
M
...
M
...
...
M
C
C
...
...
...
Sweden
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Switzerland
C
C
C
P
C
C
C
...
P
P
P
P
C
C
C
...
P
C
C
Syria
M
M
M
M
M
M
M
M
M
...
P
M
M
M
M
M
P
P
...
2
Tajikistan
M
M
M
...
D
...
...
...
M
...
...
D
D
D
D
...
...
...
...
1
Tanzania
M
M
M
M
C
C
...
C
M
...
C
C
...
...
...
...
...
C
M
TFYR
Macedonia
M
M
M
M
C
C
M
...
M
...
C
C
C
...
...
...
...
C
...
Thailand
P
...
M
...
...
P
...
C
C
...
P
C
...
M
M
P
...
C
M
Togo
P
M
P
C
C
...
...
C
M
C
P
C
...
P
P
...
...
C
P
Tonga
P
P
P
...
M
M
...
...
M
...
...
M
...
...
...
...
...
P
...
Trinidad and
Tobago
M
...
P
M
C
P
M
C
P
P
M
C
M
P
...
...
...
C
M
Tunisia
M
M
M
M
C
P
C
C
M
...
C
C
...
...
C
C
C
C
...
Turkey
P
C
C
C
C
C
C
C
C
P
P
P
C
C
C
C
P
C
C
2
Turkmenistan
C
M
M
...
M
...
...
...
D
...
C
D
...
D
D
...
...
...
...
2
Tuvalu
M
M
M
...
M
...
...
M
...
...
...
...
...
...
M
...
...
M
...
Uganda
P
P
P
P
C
P
C
P
P
P
P
C
C
P
P
...
...
...
P
1
2
1
2
2004 data
pre-2004 data
R EGULATORY T ABLES
213
Local
Long Int'l WLL
services distance
2
WORLD
Data
DSL
lines
bile ing
Cable
TV
2000
Ukraine
D
M
C
...
C
...
...
...
M
...
C
C
C
C
C
...
...
...
...
United Arab
Emirates
P
P
P
P
P
P
P
P
P
P
P
M
M
...
M
M
P
P
P
United
Kingdom
C
C
C
P
C
C
C
C
P
...
C
P
C
C
C
C
C
C
...
1
United States
C
C
C
C
C
C
C
C
C
...
C
C
C
C
C
C
C
C
...
2
Uruguay
M
M
P
M
C
M
...
C
M
...
C
C
P
C
C
C
C
C
...
2
Uzbekistan
D
D
D
...
D
...
...
...
D
...
C
C
C
D
D
...
...
...
...
2
Vanuatu
M
M
M
...
M
...
...
...
...
...
...
...
...
...
...
...
...
...
...
2
Vatican
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
Venezuela
C
C
C
C
C
C
...
C
...
...
C
C
C
...
...
...
...
C
C
Viet Nam
C
C
C
C
C
C
C
C
C
...
C
C
C
...
...
...
...
C
C
Yemen
M
M
M
...
M
M
...
M
M
...
C
M
...
M
...
C
...
C
M
Zambia
M
M
M
M
P
P
...
P
M
C
P
P
...
P
P
P
...
P
C
Zimbabwe
C
P
P
...
C
C
...
C
C
...
C
...
...
...
...
...
...
C
M
1
1
2
2004 data
pre-2004 data
214
R EGULATORY T ABLES
GLOSSARY OF TERMS *
The following definitions are included to assist the readers of this report. They are adapted from non-definitive reference
sources and are not intended to replace or contradict the terms and meanings used by each ITU Member State in its national laws
and regulations or in international agreements.
3G:
Third-generation mobile network or service.
Generic name for mobile network/service
based on the IMT-2000 family of global
standards.
802.20:
Refers to IEEE 802.20, a new standard that
is being developed for mobile broad band
wireless access.
AAS:
Adaptive Array System. A system that is
designed to enhance the detection and
reception of certain desired radio signals.
Access BPL:
A system that uses electrical distribution
lines, overhead or underground, to provide
broadband Internet access to homes and
businesses.
Active optical
network:
A network in which the passive splitting
point is replaced with Optical Line Distribution unit which is a powered unit making it
possible to have higher bit rate on individual
routes over longer distances than passive
optical network.
ADSL:
Asymmetric digital subscriber line. A technology that enables high-speed data services to
be delivered over twisted pair copper cable,
typically with a download speed in excess
of 256 kbit/s, but with a lower upload speed.
Corresponds to ITU Recommendation
(standard) ITU-T G.992.1.
ADSL2:
Asymmetric Digital Subscriber Line 2 (ITU-T
G.992.3 and ITU-T G.992.4). A sequel
to the original ITU Recommendation. It
allows increased line speeds, new powersaving elements, and extends the reach of
the original ADSL specification.
ADSL2+:
G LOSSARY
Asymmetric digital subscriber line 2 plus (ITU-T
G.992.5). This revised version of ADSL2
enables increased speeds by increasing the
frequencies used on the copper line.
OF
T ERMS
Analogue:
Transmission of voice and images using
electrical signals. Analogue mobile cellular
systems include AMPS, NMT and TACS.
Analogue
network:
A telecommunication network in which
information is conveyed as a continuously
varying electronic signal (see also Digital
network).
ASP:
Application service provider. Provider of a
service that allows users to run applications
remotely from a server rather than having
the actual programs installed on their computers. This allows for higher power applications to run on small or basic terminals.
ATM:
Asynchronous transfer mode. A transmission
mode in which the information is organized
into cells; it is asynchronous in the sense
that the recurrence of cells from an individual user is not necessarily periodic.
Bandwidth:
The range of frequencies available to be
occupied by signals. In analogue systems
it is measured in terms of Hertz (Hz) and
in digital systems in bits per second (bit/s).
The higher the bandwidth, the greater the
amount of information that can be transmitted in a given time. High bandwidth channels are referred to as “broadband” which
typically means 1.5-2.0 Mbit/s or higher.
Baseband:
Refers to transmitting on only a single channel at any one time.
Base station:
A radio transmitter/receiver and antenna
used in the mobile cellular network. It
maintains communications with cellular
telephones within a given cell and transfers
mobile traffic to other base stations and the
fixed telephone network.
215
Bent pipe star A satellite system topology characterized by
topology:
a large gateway earth station that transmits
one or more high-data-rate, forward-link
broadcasts to a large number of small user
terminals. These broadcasts contain address
information that allows each user terminal
to select those transmissions intended for it.
In the return direction, the remote user terminals transmit in bursts at low-to-medium
data rates to the gateway.
Broadband
Encompasses either mobile or fixed access
Wireless
technologies that provide connections at
Access (BWA): speeds higher than the primary rate (for
example, 2 Mbit/s).
Browser:
Application that retrieves WWW documents
specified by URLs from an HTTP server on
the Internet. Displays the retrieved documents according to the Hyptertext Markup
Language (HTML).
Bent pipe
point-topoint:
This topology calls for a dedicated duplex
connection, set up between a large gateway
earth station and a single user terminal.
Burstiness:
Bit (binary
digit):
A bit is the primary unit of electronic, digital
data. Written in base-2, binary language as a
“1” or a “0”.
Technical jargon used to describe a high
peak-to-average rate of packets as they are
received over the network. There is no
unique mathematical definition of “burstiness”, but a traffic stream is considered to be
more “bursty” than another if its packets are
more clumped together.
Byte:
(1) A set of bits that represent a single character. A byte is composed of 8 bits. (2) A bit
string that is operated upon as a unit and the
site of which is independent of redundancy
or framing techniques.
Bit/s:
Bits per second. Measurement of the transmission speed of units of data (bits) over a
network. Also kbit/s: kilobits (1 000) per
second; Mbit/s: megabits (1 000 000) per
second, and Gbit/s: Gigabits (1 000 000 000)
per second.
Blog:
Blog is short for weblog. A weblog is a journal (or newsletter) that is frequently updated
and intended for general public consumption.
Bluetooth:
Botnets:
Broadband:
A radio technology that enables the transmission of signals over short distances
between mobile phones, computers and
other devices. It is typically used to replace
cable.
A jargon term for a collection of software
robots, or bots, which run autonomously. A
botnet’s originator can control the group
remotely, usually through a means such as
IRC, and usually for nefarious purposes.
Although there exist various definitions of
broadband that have assigned a minimum
data rate to the term, it may be defined
as transmission capacity with sufficient
bandwidth to permit combined provision
of voice, data and video, with no lower
limit. Effectively, broadband is implemented
mainly through ADSL, cable modem or
wireless LAN (WLAN) services.
Broadband
corDECT:
A broadband wireless local loop standard
developed in India by Indian Institute of
Technology.
Broadband
Over Power
Line (BPL):
A wireline technology that is able to use the
current electricity networks for data and
voice transmission.
216
Cable modem: A technology that allows high-speed interactive services, including Internet access, to be
delivered over a cable TV network.
Cable
Television
(CATV):
A system for delivery of television video and
audio content via a wired network, employing either co-axial cable or fibre.
Calling Party
Pays (CPP):
Billing option whereby the person making
the call is charged. By contrast, in a “receiving party pays” (RPP) system, the individual
that receives the call pays all charges for that
call.
Cellular:
A mobile telephone service provided by
a network of base stations, each of which
covers one geographic cell within the total
cellular system service area.
CAGR:
Compound annual growth rate. See the Technical Notes.
CDMA:
Code division multiple access. A technology for
digital transmission of radio signals based
on spread spectrum techniques where each
voice or data call uses the whole radio band
and is assigned a unique code.
CDMA2000:
Code division multiple access 2000. A thirdgeneration digital cellular standard based
on Qualcomm technology. Includes
CDMA2000 1x, 1xEV-DO (Evolution, Data
Optimized) and 1xEV-DV (Evolution, Data
and Voice). One of the IMT-2000 “family”
of standards.
G LOSSARY
OF
T ERMS
Cellular:
A mobile telephone service provided by
a network of base stations, each of which
covers one geographic cell within the total
cellular system service area.
Corporatization:
Corporatization involves legal changes
to grant the telecommunication operator
administrative and financial autonomy from
central government.
Channel:
One of a number of discrete frequency
ranges utilized by a base station to transmit
and receive information from cellular terminals (such as mobile handsets).
Coverage:
Circuitswitched
connection:
A temporary connection that is established
on request between two or more stations in
order to allow the exclusive use of that connection until it is released. At present, most
voice networks are based on circuit-switching, whereas the Internet is packet-based.
See also Packet-based.
Refers to the range of a mobile cellular
network, measured in terms of geographic
coverage (the percentage of the territorial
area covered by mobile cellular) or population
coverage (the percentage of the population
within range of a mobile cellular network).
CSMA:
Carrier Sense Multiple Access. A network protocol in which a node verifies the absence of
other traffic before transmitting on a shared
physical medium, such as an electrical bus,
or a band of electromagnetic spectrum.
CMTS:
Cable Modem Termination System. Equipment
typically found in a cable company’s headend and is used to provide high speed data
services, such as Cable Internet or Voice
over IP, to cable subscribers.
Distributed
Denial of
Service
(DDoS):
Collocation:
Facility-sharing in which the incumbent
operator houses communications equipment
of competitive operators to facilitate connectivity to end users.
An attack on a computer system or network
that causes a loss of service to users, typically
the loss of network connectivity and services
by consuming the bandwidth of the victim
network or overloading the computational
resources of the victim system through a
system of computers, which are usually
zombie computers compromised by viruses
or Trojan horse programs.
Competition:
Refers to introducing competition among
national service suppliers and/or foreign
suppliers without any limitations. In the
case of mobile cellular, the number of licensees is dependent on spectrum availability.
Therefore, all countries allowing more than
one operator have been listed in this report
as “competitive”.
Digital:
Representation of voice or other information using digits 0 and 1. The digits are
transmitted as a series of pulses. Digital
networks allow for higher capacity, greater
functionality and improved quality.
Digital
network:
A telecommunication network in which
information is converted into a series of
distinct electronic pulses and then transmitted as a digital bit stream (see also Analogue
network).
DOCSIS:
Data over cable systems interface specifications
(ITU-T J.112). An ITU Recommendation
for cable modems. It specifies modulation
schemes and the protocol for exchanging
bi-directional signals over cable.
DOCSIS2:
Data over cable systems interface specifications 2
(ITU-T J.122). The newest, revised version
of DOCSIS, approved at the end of 2002.
Competitive
Local
Exchange
Carrier
(CLEC):
A network operator or carrier–often a new
market entrant--that provides local telephony in competition with the incumbent
carrier.
Condominium A network model where a group of endFibre Build:
users band together to install strands of fibre
optic cable to an ISP at the same time. At
completion, the end-users are each given
separate strands of fibre for their own usage.
Connectivity:
The capability to provide, to end-users, connections to the Internet or other communication networks.
Convergence:
A term used to describe a variety of technological and market trends involving the
blurring of previously distinct lines between
market segments such as cable television,
telephony and Internet access, all of which
can now be provided through a variety of
different network platforms.
G LOSSARY
OF
T ERMS
Domain name: The registered name of an individual or
organization eligible to use the Internet.
Domain names have at least two parts and
each part is separated by a dot (point). The
name to the left of the dot is unique for
each top-level domain name, which is the
name that appears to the right of the dot.
For instance, The International Telecommunication Union’s domain name is itu.int.
“ITU” is a unique name within the gTLD
“int”.
217
DSL:
Digital subscriber line. See also ADSL, ADSL2,
ADSL2+, SHDSL, SDSL, VDSL and
xDSL.
E-mail:
Electronic mail. The exchange of electronic
messages between geographically dispersed
locations.
DSLAM:
Digital subscriber line access multiplexer. A
device, located at the central office of a DSL
provider, that separates and routes the voicefrequency signals and data traffic on a DSL
line.
End-user:
The individual or organization that originates or is the final recipient of information
carried over a network (i.e. the consumer).
Endrun:
A fibre optic infrastructure that provides a
dedicated fibre optic cable directly to each
user’s premise rather than several premises
optically splitting off one line. See PON.
ENUM:
Standard adopted by Internet Engineering
Task Force (IETF), which uses the domain
name system (DNS) to map telephone
numbers to Web addresses or uniform
resource locators (URL). The long-term
goal of the ENUM standard is to provide
a single number to replace the multiple
numbers and addresses for users’ fixed lines,
mobile lines, and e-mail addresses.
EPOP:
Expanding point of profitability. A network
topography where the network expands
incrementally to unserved areas as they
become profitable to operators. Newly connected areas can then be used as backbones
to more remote areas as they eventually
become profitable to providers.
Ethernet:
A protocol for interconnecting computers and peripheral devices at high speed.
Recently Gigabit Ethernet has become available which enables speeds up to 1 Gbit/s.
Ethernet can run on several types of wiring
including: twisted pair, coaxial, and even
fibre optic cable.
Ex-ante and
ex-post
regulation:
Ex-ante regulation involves setting specific
rules and restrictions to prevent anti-competitive or otherwise undesirable market
activity by carriers before it occurs; ex-post
regulation, by contrast, calls for setting few
or no specific rules in advance, but applying
corrective measures and punishments if and
when transgressions do occur.
FDMA:
Frequency division multiple access. A cellular
technology that has been used in the firstgeneration analogue systems (i.e. NMT,
AMPS, and TACS).
Fixed line:
A physical line connecting the subscriber to
the telephone exchange. Typically, fixed-line
network is used to refer to the PSTN (see
below) to distinguish it from mobile networks.
DSP:
Digital signal processing. The study of signals
in a digital representation and the processing
methods of these signals
DVB-RCS:
Digital Video Broadcasting, with a Return
Channel via Satellite. An open standard for
digital television maintained by the DVB
Project, an industry consortium with more
than 270 members, and published by a Joint
Technical Committee (JTC) of European Telecommunications Standards Institute (ETSI),
European Committee for Electrotechnical
Standardization (CENELEC) and European
Broadcasting Union (EBU).
Dynamic
frequency
selection
(DFS):
It detects the presence of a primary service
and switches the WLAN to a clear frequency.
E.164:
An ITU-T recommendation which defines
the international public telecommunication
numbering plan used in the PSTN and
some other data networks.
E-commerce:
Electronic commerce. Term used to describe
transactions that take place online where the
buyer and seller are remote from each other.
ECS:
Electronic Communication Service. Services
provided for remuneration and consisting wholly or mainly in the conveyance
of signals on Electronic Communications
Networks
EDGE:
Enhanced Data rates for GSM Evolution. It acts
as an enhancement to 2G and 2.5G General
Packet Radio Service (GPRS) networks. This
technology works in TDMA and GSM networks. EDGE (also known as EGPRS) is a
superset to GPRS and can function on any
network with GPRS deployed on it, provided the carrier implements the necessary
upgrades. EDGE provides Enhanced GPRS
(EGPRS), which can be used for any packet
switched applications such as an Internet
connection. High-speed data applications
such as video services and other multimedia
benefit from EGPRS’ increased data capacity.
218
G LOSSARY
OF
T ERMS
FPGA:
Field programmable gate array. A semiconductor device containing programmable logic
components and programmable interconnects. The programmable logic components can be programmed to duplicate the
functionality of basic logic gates (such as
AND, OR, XOR, NOT) or more complex
combinatorial functions such as decoders or
simple math functions.
FTTH:
Fibre to the home. A high-speed fibre optic,
Internet connection that terminates at a residence. See FTTx.
FTTx:
Fibre to the x, where x is a home (FTTH),
building (FTTB), curb (FTTC), or neighbourhood (FTTN). These terms are used
to describe the reach of an optical fibre
network.
Firewall:
Software or hardware that controls access in
and out of a network. Firewalls can be dedicated computers that act as the intermediary
between a business network and the Internet, or can be software tools that help individual computers control which programs
are allowed access to the Internet.
First Mile:
A topology in which the user or a local
service provider – or perhaps even an apartment building company – owns the access
network and connects to service providers
using its own upstream links.
Fixed line:
A physical line connecting the subscriber to
the telephone exchange. Typically, fixed-line
network is used to refer to the PSTN (see
below) to distinguish it from mobile networks.
Frequency:
The rate at which an electrical current alternates, usually measured in Hertz (see Hz).
It is also used to refer to a location on the
radio frequency spectrum, such as 800, 900
or 1 800 MHz.
FrequencyDivision
Duplexing
(FDD):
A transmission mode that requires the
allocation of two frequency bands, one for
the uplink and another for the downlink. It
enables one to transmit and receive at the
same time.
FrequencyDivision
Multiplexing
(FDM):
A form of signal multiplexing where multiple baseband signals are modulated on
different frequency carrier waves and added
together to create a composite signal.
FSO:
Free space optics. A system of lasers used to
transmit data optically through the atmosphere at very high speeds. Similar to optical
fibre without the physical cable.
G LOSSARY
OF
T ERMS
Gateway:
Any mechanism for providing access to
another network. This function may or may
not include protocol conversion.
GATS:
General Agreement on Trade In Services.
GDP:
Gross domestic product. The market value of all
final goods and services produced within a
nation in a given time period.
GEO:
Geostationary earth orbit. A satellite in orbit
35 650 km above the Earth in a rotation that
mimics that of the Earth, thus appearing stationary in the sky.
GMPCS:
Global mobile personal communications by satellite.
Non-geostationary satellite systems that are
intended to provide global communication
coverage to small handheld devices.
Global System
for Mobile
communications (GSM):
European-developed digital mobile cellular
standard. For more information, see the
GSM Association website at:
http://www.gsmworld.com/index.html.
GNP:
Gross national product. The market value of
all final goods and services produced in a
nation’s economy, including goods and services produced abroad.
GNI:
Gross national income. The market value of
all final goods and services produced in
a nation’s economy, including goods and
services produced abroad. GNI in constant
prices, differs from GNP in that it also
includes a terms of trade adjustment; and
gross capital formation which includes a
third category of capital formation: net
acquisition of valuables.
GPRS:
General Packet Radio Service. It is a mobile
data service available to users of GSM
mobile phones. It is often described as
“2.5G”, that is, a technology between the
second (2G) and third (3G) generations
of mobile telephony. It provides moderate
speed data transfer, by using unused TDMA
channels in the GSM network.
GPS:
Global positioning system. Refers to a “constellation” of 24 “Navstar” satellites launched
initially by the United States Department
of Defense, that orbit the Earth and make it
possible for people with ground receivers to
pinpoint their geographic location. The location accuracy ranges from 10 to 100 metres
for most equipment. A Russian system,
GLONASS, is also available, and a European
system, Galileo, is under development.
H.323:
An umbrella recommendation from the
ITU-T, that defines the protocols to provide
audio-visual communication sessions on any
packet network.
219
Half duplex:
Half duplex refers to a communication
channel that can only handle one-way traffic at a time. In essence, each side of the
communication must wait until the other is
finished transmitting to start sending information. By contrast, full duplex communication allows for both parties to broadcast
and receive at the same time.
Hotspot:
An access point to a wireless local area network (WLAN). Hotspots are areas where
wireless data can be sent and received, and
Internet access is provided to wireless
devices. For example, a laptop computer can
be used to access the Internet in a hotspot
provided in an airport or hotel.
HAPS:
High altitude platform station. A term referring
to balloons and high altitude aircraft that can
be used to provide communication services.
See LAPS.
HDTV:
High-definition television. A new format for
television that offers far superior quality to
current NTSC, PAL, or SECAM systems.
The resolution of the picture is roughly
double previous television signals and the
pictures are displayed with a screen ratio of
16:9 as compared with most of today’s TV
screens, which have a screen ratio of 4:3.
Hertz (Hz):
The frequency measurement unit equal to
one cycle per second.
HFC:
Hybrid fibre copper. A broadband network that
utilizes fibre optic cabling to the vicinity and
then copper lines to individual users.
HiperLAN:
High-performance radio local area network.
An ETSI standard that operates at up to
54 Mbit/s in the 5 GHz RF band.
HiperLAN2:
High-performance radio LAN Type 2. Wireless
LAN (specified by ETSI/BRAN) in the 5
GHz IMS Band with a bandwidth up to 50
Mbit/s. HiperLAN2 is compatible with 3G
WLAN systems for sending and receiving
data, images, and voice communications.
HIPERMAN:
High performance radio metropolitan area network. This is a European standard aimed at
providing a broadband wireless solution for
Metropolitan Area Networks.
Hotspot:
An access point to a wireless local area network (WLAN). Hotspots are areas where
wireless data can be sent and received, and
Internet access is provided to wireless
devices. For example, a laptop computer can
be used to access the Internet in a hotspot
provided in an airport or hotel.
220
HSDPA:
High-Speed Downlink Packet Access. This is a
new mobile telephony protocol. Also called
3.5G (or “3½G”). High Speed Downlink
Packet Access is a packet-based data service
with data transmission up to 8-10 Mbit/s
(and 20 Mbit/s for MIMO systems) over a
5 MHz bandwidth in W-CDMA downlink.
HSDPA implementation includes Adaptive
Modulation and Coding (AMC), MultipleInput Multiple-Output (MIMO), Hybrid
Automatic Repeat Request (HARQ), fast
scheduling, fast cell search, and advanced
receiver design.
Hybrid
Fibre/Coaxial
(HFC):
A telecommunications industry term for a
network that incorporates both optical fibre
along with coaxial cable to create a broadband network.
Hz:
Hertz. The frequency measurement unit
equal to one cycle per second.
IMS:
IP Multimedia Subsystem. A standardized
Next Generation Networking (NGN) architecture for telecom operators that want to
provide mobile and fixed multimedia services. It uses a Voice-over-IP (VoIP) implementation based on a 3GPP standardized
implementation of SIP, and runs over the
standard Internet Protocol (IP). Existing
phone systems (both packet-switched and
circuit-switched) are supported.
IMT-2000:
International Mobile Telecommunications-2000.
Third-generation (3G) “family” of mobile
cellular standards approved by ITU. For
more information see the website at:
http://www.itu.int/imt.
Incumbent:
The major network provider in a particular country, often a former State-owned
monopoly.
In-house BPL: A home networking technology that uses
the transmission standards developed by the
HomePlug Alliance. Products for in-home
networking use the electric outlets in your
home (or office).
Instant
messaging
(IM):
Refers to programs such as AOL Instant
Messenger and ICQ that allow users to
exchange messages with other users over the
Internet with a maximum delay of one or
two seconds at peak times.
G LOSSARY
OF
T ERMS
Interconnection:
The physical connection of separate telephone networks to allow users of those
networks to communicate with each other.
Interconnection ensures interoperability of
services and increases end users’ choice of
network operators and service providers.
ITU:
International Telecommunication Union. The
United Nations specialized agency for telecommunications. See http://www.itu.int/.
Internet
Exchange
Point (IXP):
A central location where multiple Internet
Service Providers can interconnect their networks and exchange IP traffic.
Interconnection
Charge:
The charge–typically including a per-minute
fee–that network operators levy on one
another to provide interconnection.
JPEG:
Joint photographic expert group compression
standard. Standard for the compression and
coding of still images.
Internet:
Interconnected global networks that use the
Internet protocol (see IP).
LAN:
Internet
backbone:
The high-speed, high capacity lines or series
of connections that form a major pathway
and carry aggregated traffic within the Internet.
Internet
content
provider:
A person or organization that provides information via the Internet, either with a price
or free of charge.
Local area network. A computer network that
spans a relatively small area. Most LANs are
confined to a single building or group of
buildings. However, one LAN can be connected to other LANs over any distance via
telephone lines and radio waves. A system of
LANs connected in this way is called a widearea network (WAN). See also WLAN.
LAPS:
IP:
Internet protocol. The dominant network layer
protocol used with the TCP/IP protocol
suite.
Low altitude platform station. A system usually
consisting of balloons that provides wireless
communication services over a wide area.
Similar to HAPS but the altitudes are lower.
Last Mile:
IP telephony:
Internet protocol telephony. IP telephony is
used as a generic term for the conveyance
of voice, fax and related services, partially
or wholly over packet-based, IP-based networks. See also VoIP and Voice over
broadband.
The topology denotes the operator’s ownership of the access network.
Layered
Architecture:
The concept of layered network architecture divides a network at any specific point
into layers, each of which adds value to the
physical medium of communication.
LBS:
Location-based services. LBS make use of
information on the location of a mobile
device and user, and can exploit a number
of technologies for the geographic location
of a user. Some of these technologies are
embedded in the networks and others in the
handsets themselves. Location capability is
already available to some level of accuracy
(approx. 150 m) for most users of cellular
networks. Increased accuracy can become
available through location technologies such
as GPS. See GPS.
Leased line:
A point-to-point communication channel
or circuit that is committed by the network
operator to the exclusive use of an individual subscriber. Under national law, leased
lines may or may not be permitted to interconnect with the public switched network.
LEO:
Low Earth orbit. A term that refers to satellite
orbits between 650 km and 2 600 km above
the Earth. A LEO satellite is only in view for
a few minutes and rotates the Earth every
few hours. See GEO.
Licensing:
An administrative procedure for selecting
operators and awarding franchises for the
operation of particular telecommunication
services, for instance cellular radio.
IPR:
Intellectual property rights. Copyrights, patents
and trademarks giving creators the right to
prevent others from using their inventions,
designs or other creations. The ultimate aim
is to act as an incentive to encourage the
development of new technology and creations which will eventually be available to all.
The main international agreements are the
World Intellectual Property Organization’s
(WIPO) Paris Convention for the Protection of
Industrial Property (patents, industrial designs,
etc.), the Berne Convention for the Protection
of Literary and Artistic Works (copyright), and
the World Trade Organization’s (WTO)
Agreement on Trade-Related Aspects of Intellectual
Property Rights (TRIPS).
ISDN:
Integrated services digital network. A digital
switched network, supporting transmission
of voice, data and images over conventional
telephone lines.
ISP:
G LOSSARY
Internet service provider. ISPs provide endusers access to the Internet. Internet Access
Providers (IAPs) may also provide access to
other ISPs. ISPs may offer their own proprietary content and access to online services
such as e-mail.
OF
T ERMS
221
Line Sharing:
A form of network unbundling that allows a
competitive service provider to offer ADSL
using the high-frequency portion of a local
loop at the same time that an ILEC continues to offer standard switched voice service
over the low-frequency portion of the same
loop.
MP3:
MPEG-1 Audio Layer-3 (MPEG stands for
Moving Pictures Experts Group). A standard
technology and format for compression of a
sound sequence into a very small file (about
one-twelfth the size of the original file)
while preserving the original level of sound
quality when it is played.
LLU:
Local loop unbundling. The process of requiring incumbent operators to open the last
mile of their legacy networks to competitors.
Similar reference to ULL (unbundled local
loop).
Multimedia:
The presentation of more than one medium,
typically images (moving or still), sound and
text in an interactive environment. Multimedia requires a significant amount of data
transfer and bandwidth, and it invariably
requires computational facilities.
Local loop:
The system used to connect the subscriber
to the nearest switch. It generally consists of
a pair of copper wires, but may also employ
fibre-optic or wireless technologies.
National
Regulatory
Authority
(NRA):
The regulatory agency or official at the
central or federal government level that is
charged with implementing and enforcing
telecommunication rules and regulations.
Network
Unbundling:
Providing access to, or making available,
some or all of the disaggregated elements
and/or functions of a telephone network–
usually the local portion of the network–for
interconnecting carriers to use in serving
their own customers.
Network
Topology:
The pattern of links connecting pairs of
nodes of a network.
Next generation network
(NGN):
These are packet-based networks in which
service-related functions are independent
from underlying transport-related technologies. They are able to provide telecommunication services and make use of multiple
broadband transport technologies. For a
more detailed definition, see http://www.itu.
int/ITU-T/ngn
Node:
A point of connection to a network. A
switching node is a point at which switching
occurs.
Number
portability:
The ability of a customer to transfer an
account from one service provider to
another without requiring a change in
number. Other forms of portability allow
end users to change residence or subscribe
to a new form of service (e.g., ISDN) while
retaining the same telephone number for
their main telephone line.
OFDM:
Orthogonal frequency division multiplexing. A
method of digital modulation in which a
signal is split into several narrowband channels at different frequencies in order to
minimize interference among channels that
are close in frequency. OFDM is used in
European digital audio broadcast services,
and also in wireless LANs.
Long Run
The added or extra cost entailed in providIncremental
ing a service, over the long term.
Costs (LRIC):
Long Run
Average
Incremental
Costs (LRAIC):
A costing model based on LRIC analysis, in
which the total traffic costs for both interconnecting carriers are divided by the total
demand, rather than assigning unique costs
to each operator.
Main
Telephone line connecting a subscriber to
telephone line: the telephone exchange equipment. This
term is synonymous with the term fixed line
used in this report.
Market
The shortfall between commercially availefficiency gap: able access and universal access.
Mesh network: A way to route data, voice and instructions
between nodes. It allows for continuous
connections and reconfiguration around
blocked paths by “hopping” from node to
node until a connection can be established.
Mobile:
As used in this report, the term refers to
mobile cellular systems and to mobile
phones.
Mobile virtual
network
operator
(MVNO):
A company that does not own a licensed frequency spectrum, but resells wireless services under their own brand name, using the
network of another mobile phone operator.
Multimedia:
The presentation of more than one medium,
typically images (moving or still), sound and
text in an interactive environment. Multimedia requires a significant amount of data
transfer and bandwidth, and it invariably
requires computational facilities.
222
G LOSSARY
OF
T ERMS
On-Board
Processor
(OBP)
Switching:
In this topology, the satellite rather than the
gateway is the central node in a star network.
The satellite is connected to the gateway by
one or more high-data-rate trunks. The onboard processor de-multiplexes the uplink
trunk into several downlinks for different
geographical areas, usually determined by
the footprint pattern. The forward downlinks contain messages for large numbers
of user terminals, and the destinations are
identified by message headers. In the return
channel, the uplink transmissions from user
terminals in one or more cells are multiplexed onto a downlink trunk to the gateway.
One DSL for Universal Service (See UDSL/UniDSL)
OperatorA network access model in which the passive
neutral model: access network infrastructure is often owned
by housing companies, condominiums or
tenant organizations and, in some cases,
by municipalities. In this model, access
networks are connected to a shared access
network backbone. Any service provider can
then connect its network gateway and offer
services using the access network.
OPGW:
Optical power ground wire. A special ground
wire with a fibre cable in the core.
Optical
network
terminal
(ONT):
The equipment used to terminate the fiber.
P2P:
Peer to peer. P2P refers to networks that
facilitate direct connections among individual nodes rather than through a centralized server. However, many famous P2P
networks, such as “Napster”, actually relied
on a central server to connect users. Other
networks (such as “Gnutella”) offer true
peer-to-peer, decentralized connections.
PABX:
Private Automatic Branch eXchange. A telephone exchange that is owned by a private
business, as opposed to one owned by a
common carrier or by a telephone company.
Packet:
Block or grouping of data that is treated as
a single unit within a communication network.
Packet-based:
Message-delivery technique in which packets are relayed through stations in a network.
See also Circuit-switched connection.
G LOSSARY
OF
T ERMS
PAN:
Personal area network. For the purposes of this
report, a PAN is referred to as the interconnection of information technology devices
within the range of an individual person,
typically within a radius of 10 metres. For
example, a person travelling with a laptop, a
personal digital assistant (PDA), and a portable printer could interconnect these devices
through a wireless connection, without the
need for physical wiring. Conceptually, the
difference between a PAN and a wireless
LAN is that the former tends to be centered
around one person while the latter has a
greater range of wireless connectivity, typically serving multiple users.
PDA:
Personal digital assistant. A generic term for
handheld devices that combine computing
and possibly communication functions.
Peering:
The exchange of routing announcements
between two Internet Service Providers for
the purpose of ensuring that traffic from
the first can reach customers of the second,
and vice-versa. Peering takes place predominantly at IXPs and usually is offered either
without charge or subject to mutually agreed
commercial arrangements.
Penetration:
A measurement of access to telecommunications, normally calculated by dividing the
number of subscribers to a particular service
by the population and multiplying by 100.
Also referred to as teledensity (for fixed-line
networks) or mobile density (for cellular
ones), or total teledensity (fixed and mobile
combined).
Personal
Communication Services
(PCS):
In the United States, this refers to digital
mobile networks using the 1 900 MHz frequency. In other countries, it refers to digital
mobile networks using the 1 800 MHz frequency (See DCS-1800). The term Personal
Communications Network (PCN) is also used.
Pervasive
computing:
A concept which describes a situation in
which computing capability is embedded
into numerous different devices around
the home or office (e.g. fridges, washing
machines, cars, etc.). Also referred to as
ubiquitous computing. Pervasive communications
implies that the microchips in these devices
are also able to communicate, for instance
their location and status.
Phishing:
The fraudulent practice of disguising spam
as legitimate email in an attempt to coax
recipients into revealing private financial
data.
223
PLC:
PON:
Power line communications. A communication
network that uses existing power lines to
send a receive data by using electrical signals
as the carrier. Power flows on the line at
50-60 Hz while data is sent in the 1 MHz
range.
PTO:
Public telecommunication operator. A provider of
telecommunications infrastructure and services to the general public (“public” refers to
the customer base). Also referred to as an
operator, service provider, carrier or “telco”.
QoS:
Passive optical network. A type of full passive
wave division multiplexing (WDM) network
that allows multiple locations to connect to
one optical fibre strand (or wavelength) by
using optical splitters to break up the wavelength of light into allocated time slots for
each user. See Endrun and WDM.
Quality of service. A measure of network
performance that reflects the quality and
reliability of a connection. QoS can indicate
a data traffic policy that guarantees certain
amounts of bandwidth at any given time, or
can involve traffic shaping that assigns varying bandwidth to different applications.
RLAN:
Radio local area network. See WLAN.
RFID:
Radio-frequency identification. A system of radio
tagging that provides identification data for
goods in order to make them traceable. Typically used by manufacturers to make goods
such as clothing items traceable without
having to read bar code data for individual
items.
SCADA:
Supervision, control and data acquisition. The
term refers to a large-scale, distributed
measurement (and control) system.
SDR:
Software defined radio. A radio communication
system which uses software for the modulation and demodulation of radio signals.
SDSL:
Symmetrical DSL. A proprietary North
American DSL standard. However, the term
SDSL is often also used to describe SHDSL.
Secondgeneration
mobile (2G):
A general term for second generation mobile
networks, for example GSM.
Server:
(1) A host computer on a network that sends
stored information in response to requests
or queries. (2) The term server is also used
to refer to the software that makes the process of serving information possible.
Session
Initiation
Protocol
(SIP):
A protocol developed by the IETF
MMUSIC Working Group and proposed
standard for initiating, modifying, and
terminating an interactive user session that
involves multimedia elements such as video,
voice, instant messaging, online games, and
virtual reality. In November 2000, SIP was
accepted as a 3GPP signalling protocol and
permanent element of the IMS architecture.
It is one of the leading signalling protocols
for Voice over IP, along with H.323.
SHDSL:
Single pair high-speed DSL. The informal
name for ITU-T Recommendation G.991.2
that offers high-speed, symmetrical connectivity over a twisted copper pair.
Port 25:
The traditional TCP port used by the
Simple Mail Transfer Protocol.
Portal:
Although an evolving concept, the term
portal commonly refers to the starting point,
or a gateway through which users navigate
the World Wide Web, gaining access to a
wide range of resources and services, such
as e-mail, forums, search engines, and shopping malls.
PPP:
Protocol:
Purchasing power parity. An exchange rate that
reflects how many goods and services can
be purchased within a country taking into
account different price levels and cost of
living across countries.
A set of formal rules and specifications
describing how to transmit data, especially
across a network.
Primary rate
connection
(PRI):
One kind of service for ISDN. It consists of
23 B-channels and one 64 kbit/s D-channel
in the United States or 30 B-channels and 1
D-channel in Europe.
Private
network:
A network based on leased lines or other
facilities, which are used to provide telecommunication services within an organization
or within a closed user group as a complement or a substitute to the public network.
Private
ownership/
Privatization:
The transfer of control of ownership of a
state enterprise to private parties, generally
by organizing the enterprise as a share company and selling shares to investors. More
generally, the term is sometimes used to
refer to a wide range of modalities whereby
business is opened to private enterprise and
investment.
PSTN:
Public switched telephone network. The public
telephone network that delivers fixed telephone service.
224
G LOSSARY
OF
T ERMS
Short Message A service available on digital mobile cellular
Service (SMS): networks and even landline telephones, typically enabling end users to send and receive
messages with up to 160 characters.
SkyWrap:
A process, where there already are power
lines, that wraps fibre around the transmission lines.
SMTP:
Simple Mail Transfer Protocol. The de facto
standard for email transmission across the
Internet.
Softswitch:
A type of telephone switch that uses software running on a computer system to carry
out the work that used to be carried out by
hardware.
Spam:
Synchronous
Optical
Network
(SONET):
A standard for communicating digital
information using lasers or light emitting
diodes (LEDs) over optical fiber as defined
by GR-253-CORE from Telcordia. It was
developed to replace the PDH system for
transporting large amounts of telephone and
data traffic and to allow for interoperability
between equipment from different vendors.
TCP:
Transmission control protocol. A transport layer
protocol that offers connection-oriented,
reliable stream services between two hosts.
This is the primary transport protocol used
by TCP/IP applications.
TDMA:
Time Division Multiple Access. It is a technology for shared medium (usually radio)
networks. It allows several users to share the
same frequency by dividing it into different
time slots. The users transmit in rapid succession, one after the other, each using their
own timeslot. This allows multiple users to
share the same transmission medium (e.g.
radio frequency) whilst using only the part
of its bandwidth they require.
TD-SCDMA:
Time Division Synchronous Code-Division Multiple Access. A 3G mobile telecommunications
standard, being pursued in China by the
Chinese Academy of Telecommunications
Technology (CATT), Datang and Siemens
AG, in an attempt to develop home-grown
technology and not be “dependent on Western technology.” TD-SCDMA uses TDD,
in contrast to the FDD scheme used by
W-CDMA.
Unsolicited commercial email, some of
which may contain computer viruses or
worms, fraudulent consumer scams or
offensive content.
Spam Zombie: (see Zombie)
Spectral
Efficiency:
A measure of the performance of encoding
methods that code information as variations
in an analogue signal.
Spectrum:
The radio frequency spectrum of hertzian
waves used as a transmission medium for
cellular radio, radiopaging, satellite communication, over-the air broadcasting and other
services.
Spreadspectrum
technology:
A radio technique that continuously alters
its transmission pattern either by constantly
changing carrier frequencies or by constantly changing the data pattern.
SPIM:
Spam over Instant Messenger. An unsolicited
message made using instant messenger over
the internet.
SPIT:
Spam over Internet Telephony. An unsolicited
telephone call made using IP telephony over
the internet.
Switch:
Part of a mobile or fixed telephone system
that routes telephone calls or data to their
destination.
Synchronous
Digital
Hierarchy
(SDH):
A standard developed by ITU (G.707 and
its extension G.708) that is built on experience in the development of SONET. Both
SDH and SONET are widely used today;
SONET in the U.S. and Canada, SDH in
the rest of the world. SDH is growing in
popularity and is currently the main concern
with SONET now being considered as the
variation.
G LOSSARY
OF
T ERMS
Time-Division Enables use of the same frequency for
Duplexing
uplink and downlink of a transmission by
(TDD):
allocating discrete, short-duration time slots
to the two links.
Teledensity:
Number of main telephone lines per 100
inhabitants. See Penetration.
ThirdA general term for the next generation of
Generation
broadband digital mobile cellular systems,
Mobile (3G): which will have expanded broadband capabilities for mobile data applications. See
IMT-2000.
Total
Sum of the number of fixed lines and
teledensity:
mobile phone subscribers per 100 inhabitants. See Penetration.
TPC:
Transmit power control. A technical mechanism
used within some networking devices in
order to prevent too much unwanted interference between different wireless networks.
Traffic
Traffic exchange points are used by operators
Exchange
to exchange traffic through peering directly
Point:
between service networks rather than indirectly, via transit through their upstream
providers.
225
Triple Play:
A term refers to the bundling of voice, video
and data services.
True access
gap:
The shortfall between market-based regulatory measures and universal access.
UASL:
Unified Access Services Licensing. A licensing
framework in India that gives the licensee
freedom to offer both fixed and mobile services using any technology.
Ubiquitous
computing:
UDSL/
UniDSL:
A term that reflects the view that future
communication networks will allow seamless access to data, regardless of where the
user is. See Pervasive computing.
A new variant of DSL, integrating all earlier DSL variants. It promises aggregated
bit rates of up to 200 Mbit/s, including
100 Mbit/s symmetrical connections. UniDSL would require a fibre backbone infrastructure and would use only the part of the
existing subscriber line closest to the user
premises.
VDSL:
Very-high-data-rate digital subscriber line.
(ITU-T G.993.1). The fastest version of
DSL that can handle speeds up to 52 Mbit/s
over very short distances. Often used to
branch out from fibre connections inside
apartment buildings.
Voice over
broadband:
A method of making voice calls over a
broadband connection. The calls can be
either made via a computer or through
traditional phones connected to voice over
broadband equipment. See also IP telephony
and VoIP.
VoIP:
Voice over IP. A generic term used to describe
the techniques used to carry voice traffic
over IP (see also IP telephony and Voice over
broadband).
VPN:
Virtual private network. A method of encrypting a connection over the Internet. VPNs
are used extensively in business to allow
employees to access private networks at
the office from remote locations. VPNs are
especially useful for sending sensitive data.
VSAT:
Very Small Aperture Terminal. A 2-way satellite ground station with a dish antenna that
is smaller than 3 meters, as compared to
around 10 meters for other types of satellite
dishes.
WAN:
Wide area network. WAN refers to a network
that connects computers over long distances.
Wavelength
division
multiplexing:
A technology which multiplexes multiple
optical carrier signals on a single optical
fibre by using different wavelengths (colours) of laser light to carry different signals.
ULL:
Unbundled local loop. See LLU.
UMTS:
Universal mobile telecommunications system. The
European term for third-generation mobile
cellular systems or IMT-2000 based on the
W-CDMA standard. For more information
see the UMTS Forum website at:
http://www.umts-forum.org/.
Universal
access:
Refers to reasonable telecommunication
access for all. Includes universal service for
those that can afford individual telephone
service and widespread provision of public
telephones within a reasonable distance of
others.
USO:
Universal service obligations. Requirements
that governments place on operators to offer
service in all areas, regardless of economic
feasibility.
W-CDMA:
Wideband code division multiple access. A thirdgeneration mobile standard under the IMT2000 banner, first deployed in Japan. Known
as UMTS in Europe. See also CDMA.
UTP:
Unshielded twisted pair. A cable with one
or more twisted copper wires bound in a
plastic sheath. It is used extensively for highspeed connections because it allows the
release of radiation that would interfere if
kept in the line with a shielded cable.
WDM:
Wave division multiplexing. Technology that
allows multiple data streams to travel simultaneously over the same fibre optic cable by
separating each stream into its own wavelength of light.
Value-added
network services (VANS):
Telecommunication services provided over
public or private networks which, in some
way, add value to the basic carriage, usually
through the application of computerized
intelligence. Examples of VANs include
reservation systems, bulletin boards, and
information services. Also known as enhanced
services.
Website/
Webpage:
A website (also known as an Internet site)
generally refers to the entire collection of
HTML files that are accessible through a
domain name. Within a website, a webpage
refers to a single HTML file, which when
viewed by a browser on the World Wide
Web could be several screen dimensions
long. A “home page” is the webpage located
at the root of an organization’s URL.
226
G LOSSARY
OF
T ERMS
Wi-Bro:
A wireless networking technology (IEE
802.16x) that will enable portable Internet access. The Republic of Korea has
announced plans to commercialize it in
2005-2006.
Wi-Fi:
Wireless fidelity. A mark of interoperability
among devices adhering to the 802.11b
specification for Wireless LANs from the
Institute of Electrical and Electronics Engineers (IEEE). However, the term Wi-Fi is
sometimes mistakenly used as a generic
term for wireless LAN.
Wi-Fi5:
Wireless fidelity 5. A mark of interoperability
among devices adhering to the 802.11a
standard at 5 MHz.
WiMAX:
Fixed wireless standard IEEE 802.16 that
allows for long-range wireless communication at 70 Mbit/s over 50 kilometres. It can
be used as a backbone Internet connection
to rural areas.
Wireless:
Generic term for mobile communication
services which do not use fixed-line networks for direct access to the subscriber.
WISP:
An ISP that employs a wireless access platform, or “wireless ISP”.
WLAN:
Wireless local area network. Also known as
Wireless LAN. A wireless network whereby
a user can connect to a local area network
(LAN) through a wireless (radio) connection, as an alternative to a wired local area
network. The most popular standard for
wireless LANs is the IEEE 802.11 series.
* Source:
WLL:
Wireless local loop. Typically a phone network
that relies on wireless technologies to provide the last kilometre connection between
the telecommunication central office and
the end-user.
Worm:
A self-contained program (usually malicious)
that can automatically propagate throughout
a network. In addition to damage caused
by the program on a user’s machine, the
programs can slow down network traffic as
all infected machines scan simultaneously to
find new hosts.
WSIS:
The United Nations World Summit on the Information Society. The first phase of WSIS took
place in Geneva (hosted by the Government
of Switzerland) from 10 to 12 December
2003, and the second phase in Tunis (hosted
by the Government of Tunisia), from 16 to
18 November 2005. For more information
see: http://www.itu.int/wsis.
WTO Agreement:
Informal terminology for the Fourth Protocol to the General Agreement on Trade in
Services (GATS). The agreement, concluded
in early 1997, included commitments by
more than 70 countries to open their markets for basic telecommunication services.
The accompanying Reference Paper, spelled
out principles for regulatory treatment of
basic telecommunication service providers,
including “major suppliers.”
xDSL:
While DSL stands for digital subscriber line,
xDSL is the general representation for various types of digital subscriber line technology, such as ADSL, SHDSL, and VDSL. See
ADSL, SHDSL, VDSL.
Zombie:
A computer attached to the Internet that has
been compromised by a cracker, a computer
virus, or a Trojan horse program.
ITU, World Telecommunication Development Report, 1999.
ITU, Challenges to the Network, Internet Development, 1999.
G LOSSARY
OF
T ERMS
227
LIST OF FIGURES
Figure 1.1:
Figure 1.2:
Figure 1.3:
Figure 1.4:
Figure 1.5:
Figure 1.6:
Figure 1.7:
Figure 1.8:
Figure 1.9:
Figure 1.10:
Figure 1.11:
Figure 1.12:
Figure 1.13:
Figure 1.14:
Figure 1.15:
Figure 1.16:
Figure 1.17:
Figure 1.18:
Figure 1.19:
Figure 1.20:
Figure 2.1:
Figure 3.1:
Figure 3.2:
Figure 3.3:
Figure 3.4:
Figure 3.5:
Figure 3.6:
Figure 3.7:
Figure 3.8:
Figure 3.9:
Figure 3.10:
Figure 3.11:
Figure 3.12:
Figure 3.13:
Figure 4.1:
Figure 5.1:
Figure 5.2:
Figure 5.3:
Figure 5.4:
Figure 5.5:
Figure 5.6:
Figure 6.1:
228
The Number of ICT Users Worldwide, 1994-2004 .........................................................2
Fixed vs. Mobile Teledensity by Region (per 100 subscribers) ......................................3
Broadband ..................................................................................................................3
Global Distribution of Internet and Broadband Subscribers, 2004 ..................................4
Top 25 Broadband Subscribers, Non-OECD Countries (2004) .......................................4
Subscriber Growth in the Top 10 Broadband Countries (2000-2004) .............................5
Broadband Platform Distribution (by Region) 2004 .......................................................6
Top 10 3G Mobile Markets Worldwide, 2005 ................................................................7
2G Standards Usage, by subscribers, world and by region (2004-5) .............................8
Privatizations, 1991-2005, World.................................................................................9
Status of Competition Worldwide, 2005 .....................................................................10
Status of Public Interconnection Agreements and Pricing Information, 2005 ................11
Countries Requiring Local Loop Unbundling, 2005 .....................................................12
Number of Countries with National and SubRegional IXPs, by region, 2005 .................13
Growth of Regulators Worldwide, 2005.......................................................................13
Separate Regulators, by Region, 2005........................................................................14
Growth of International Traffic .....................................................................................16
Are Licences Required for WLANS, 2005? ..................................................................17
Spam as Percentage of Emails Worldwide, 2003-05 ...................................................17
Spam Regulation, 2005 .............................................................................................18
WiMAX Growth, Forecast Sales, 2004-5 .....................................................................25
Comparative Prices for Mobile Data Services .............................................................34
Migration Paths from 2G to 3G ...................................................................................35
A mesh network topology versus a traditional network.................................................38
The Convergence of WLAN and 3G Technologies – A Matrix........................................39
Empirical Data for Tower, Installation, and Grounding Costs ........................................40
Linear Model of Tower, Installation, and Grounding costs (USA/Ghana) .......................40
Radio Tower Types and Heights ..................................................................................41
Mobility v. Data Rate for Popular BWA Systems ..........................................................44
Power Consumption of Some Wireless Radio Products ...............................................45
UNHCR Refugee Camps in Tanzania ...........................................................................46
The KTTC Computer Laboratory..................................................................................47
KTTC’s Eco-Friendly Power Plant ...............................................................................47
The Mtabila Camp VSAT and PV Systems ..................................................................48
A Simplified Broadband Value Chain ..........................................................................55
From Silos to Layers ..................................................................................................76
From VLF to EHF .......................................................................................................79
Changing the Technology Paradigm ...........................................................................79
Globally Harmonized Spectrum: IMT-2000 ................................................................85
BWA Coverage Areas (Circles) in Ireland .................................................................... 87
The 2.5-2.7 GHz BWA Allocation in Mauritius ............................................................89
Shades of Grey ..........................................................................................................93
L IST
OF
F IGURES
LIST OF TABLES
Table 1.1: Newly Created Regulatory Authorities, 2005 .................................................................... 15
Table 3.1: Relative Functionalities of Broadband Wireless Access Technologies ............................... 33
L IST
OF
T ABLES
229
LIST OF BOXES
Box 1.1:
Box 2.1:
Box 3.1:
Box 3.2:
Box 3.3:
Box 3.4:
Box 3.5:
Box 3.6
Box 4.1:
Box 4.2:
Box 4.3:
Box 4.4:
Box 4.5:
Box 4.6:
Box 4.7:
Box 4.8:
Box 4.9:
Box 4.10:
Box 4.11:
Box 4.12:
Box 4.13:
Box 4.14:
Box 4.15:
Box 4.16:
Box 4.17:
Box 4.18:
Box 5.1:
Box 5.2:
Box 5.3:
Box 5.4:
Box 5.5:
Box 5.6:
Box 6.1:
Box 6.2:
Box 6.3:
Box 6.4:
Box 6.5:
Box 6.6:
Box 6.7:
Box 6.8:
Box 7.1:
230
Broadband Wireless Nations ...............................................................................................7
Wi-Fi Beyond Hotspots ................................................................................................... 25
A Radio Transmission Primer ........................................................................................... 34
TD-SCDMA – A Chinese Standard ................................................................................... 35
802.16 Extensions in the Works ....................................................................................... 36
The Evolution of a “Southern” Solution ............................................................................ 36
Mesh Networks ...............................................................................................................37
Convergence of WLAN and 3G Networks .......................................................................... 39
GSR 2004 Best Practice Guidelines ............................................................................52-53
The Impact of Broadband Access – Examples in Developing Countries ............................. 55
Regulatory Methods To Boost Deployment ........................................................................57
Licensing Incentives for Network Deployment, selected examples .................................... 59
Unified Licensing Frameworks, selected examples ............................................................61
Satellite Broadband in Developing Countries, selected examples ...................................... 62
Local Loop Wholesale Options ......................................................................................... 63
Local Loop Unbundling in Poland ..................................................................................... 63
Removing Network Bottlenecks in India ............................................................................ 65
Using the Indian Rail and Gas Facilities for Backbone Connectivity .................................. 65
GSR 2003 Universal Access Best Practice Guidelines ................................................. 66-67
Korea’s KII Project ........................................................................................................... 68
Use of Universal Access Funds, selected examples .......................................................... 68
Limiting Cross-Ownership in the EU ................................................................................ 70
E-Government in Vietnam ................................................................................................ 70
The Rural-Enlaces Project in Chile ....................................................................................71
Installing Internet Centres in Southern Brazil ....................................................................71
Encouraging PC penetration, selected examples .............................................................. 72
Defining Broadband Wireless Access ...............................................................................77
Spectrum Bands for BWA ................................................................................................77
Software Defined Radios, Adaptive Array Systems and Mesh Networks ............................. 80
Eire’s Response to BWA .................................................................................................. 82
Eire’s Response to BWA, Part Two .....................................................................................87
OFTA’s Consultation on Broadband Wireless Access Licensing ........................................ 88
A VoIP Primer ................................................................................................................. 92
The VoIP Transition ...........................................................................................................97
Classification of VoIP Services....................................................................................... 100
Emergency Services under the North American Numbering Plan..................................... 103
The FCC’s Enhanced 911 Service Order1 ....................................................................... 103
SPIT: A Looming Issue ....................................................................................................107
Threats to VoIP Networks and Publicly Available Services ................................................107
Defense Mechanisms against Security Attacks ................................................................107
Australia Telecommunications Act 1997 – SECT 117 ......................................................121
L IST
OF
B OXES
ALSO AVAILABLE FROM ITU
PUBLICATIONS
Trends in Telecommunication Reform: Licensing in an Era of Convergence, 2004/05 (6th Edition) ..............................95 CHF
Trends in Telecommunication Reform: Promoting Universal Access to ICTs, 2003 (5th Edition) ..................................90 CHF
Trends in Telecommunication Reform: Effective Regulation, 2002 (4th Edition) ...........................................................90 CHF
Trends in Telecommunication Reform: Interconnection Regulation, 2000-2001 (3rd Edition) .......................................90 CHF
Trends in Telecommunication Reform: Convergence and Regulation, 1999 (2nd Edition) .............................................75 CHF
General Trends in Telecommunication Reform 1998: World Volume I...........................................................................75 CHF
General Trends in Telecommunication Reform 1998: Africa Volume II ..........................................................................65 CHF
General Trends in Telecommunication Reform 1998: Americas Volume III ...................................................................55 CHF
General Trends in Telecommunication Reform 1998: Arab States Volume IV ...............................................................45 CHF
General Trends in Telecommunication Reform 1998: Asia Pacific Volume V ................................................................60 CHF
General Trends in Telecommunication Reform 1998: Europe Volume VI .......................................................................72 CHF
Collection of five Regional reports (Volumes II-VI) ......................................................................................................297 CHF
Collection of Regional and World reports (Volumes I-VI).............................................................................................372 CHF
The Arab Book: Telecommunication Policies for the Arab Region, 2002 ......................................................................50 CHF
The Blue Book: Telecommunication Policies for the Americas, 2000 (2nd Edition) ......................................................50 CHF
Please contact the ITU Sales Service:
Tel.:
+41 22 730 5111
Fax:
+41 22 730 5194
E-mail:
[email protected]
Website: www.itu.int
© ITU 2006
International Telecommunication Union
Place des Nations
CH-1211 Geneva, Switzerland
First printing 2006
All rights reserved. No part of this publication may be reproduced, by any means whatsoever, without the prior written
permission of the International Telecommunication Union.
Denominations and classifications employed in this publication do not imply any opinion on the part of the International
Telecommunication Union concerning the legal or other status of any territory or any endorsement or acceptance of any boundary.
Where the designation “country” appears in this publication, it covers countries and territories.
ISBN 92-61-11431-8
Note: Discounts are available for all ITU Member States, Sector Members and Least Developed Countries.
U n i o n
2 020060 6
N
I
S
D
TRECONMMUNICATION
*28388*
Printed in Switzerland
Geneva, 2006
ISBN 92-61-11431-8
T R E N D S I N T E L E C O M M U N I C AT I O N R E F O R M
TELE
m
r
o
f
re
6
0
20
G
N
I
T
A
L
REGU
E
H
T
I N ADBAND
BRO
D
L
R
WO
International
Union

trends in telecommunication reform 2006

Transcription

Similar documents

Carlota Reyners Fontana - Cabinet EC VP Neelie Kroes

Atlantic Broadband going to all-digital video network

CI slide presentation - black

WHO ARE THE ECHO BOOMERS?

Northern Rivers RDA Michael Whereat – Sunshine Coast Council

CBeyond BeyondMobile

Welcome to TV and broadband from Canal Digital

get the most from your broadband to go hotspot Bill Pay

View - Southern Regional Assembly

Customer code of practice