report - Dr. George Markowsky

Transcription

National Science Foundation Workshop
ANYWHERE, ANYTIME, ANY SIZE, ANY SIGNAL:
Scalable, Remote Information Sensing
and Communication Systems Workshop
January 14-15, 2002
ACCESS Center
Arlington, VA
edited by
George Markowsky, University of Maine
David Nagel, The George Washington University
Don Mitchell, NCSA
Janet Thot-Thompson, NCSA
Tom Williams, Air Networking
Executive Summary
Just as wild animals depend on the keenness of their senses to detect threats and
opportunities, modern society requires new sensors and modes of communication to
manage threats to homeland security and the environment. Emerging technologies in
sensor design and wireless communications have the potential to make a leap in both the
quantity and quality of critically needed data. This workshop brought together experts in
sensors, high-speed wireless communication, networking, power sources, FCC
regulations, crisis management, security, transportation infrastructure and field science to
review recent developments and discuss the potential for deployment of networked
sensors for field science and homeland security. Additionally, they worked to identify
opportunities and impediments for utilization of sensor networks in meeting the nation's
information needs.
The presenters showed how sensors and imagers, enabled by micro and other
technologies, form sensor systems that acquire information. Networks can distribute that
information over local to global and larger distances, enabling truly remote sensing.
Wireless single-link sensor systems and multiple-link networks offer features that can
include high performance, lower installation costs, reconfigurability and mobility. They
are complex engineering systems, which have widely varying characteristics and
performance. Wireless sensor nodes, sometimes called sensor clusters, tend to have a
common set of components regardless of their applications. Energy sources frequently
limit the performance and lifetime of deployed sensor nodes. There is a current need for
software to co-design and simulate the performance of sensor nodes, systems and
networks. Applications of real-time wireless sensor systems and networks are increasing
rapidly now. They include (a) determination of weather and other environmental
conditions for many reasons, including agriculture, (b) monitoring energy, fluid,
machinery, information and other systems in factories, facilities, buildings and homes, as
well as the structures themselves, (c) sensing terrestrial, marine, air and space vehicles
and associated transportation systems, (d) monitoring people and equipment for safety,
health and medical reasons, (e) sensing for security, crisis management and military
operations, and (f) diverse other applications. Ubiquitous sensing will increasingly be
part of the fabric of life in technological societies. Wireless sensor systems and networks
enable radically new integration of engineered systems and humans.
The workshop identified important problems in the design of sensor systems, such as
developing adequate power systems, that must be solved if we are to build and deploy
successful sensor systems. Besides the technical problems that were identified, a variety
of political and social issues were identified as well. In particular, the United States must
seriously rethink the way that it uses its spectrum and find ways to enable new spread
spectrum technologies to be used effectively, which will enable widespread use of
inexpensive wireless systems. Similarly, it is essential that we develop a mechanism to
get people to think seriously about security in the widest possible sense. These and many
additional recommendations are presented throughout this report, along with much
background information.
The presentations and discussions were very informative and lively. The first day
presentations may be viewed at http://homeland.maine.edu/anywhere.htm along with the
Anytime, Anywhere, Any Size and Any Signal Workshop Proceedings
page-ii
slides that were presented. This proceedings blends the presentations into a single
document, and identifies issues that need further study and development.
page-iii
Table of Contents
Table of Contents................................................................................................................iv
Chapter 1. Introduction....................................................................................................... 1
1.1. Overview Of Workshop ............................................................................................... 1
Chapter 2. Introduction To Homeland Security................................................................ 14
2.1 Introduction................................................................................................................. 14
2.2. The Big Picture........................................................................................................... 15
2.3 The Threat to Homeland Security............................................................................... 16
2.4 Transportation, Buildings and Homes......................................................................... 20
2.5 Medical Applications and Biodefense......................................................................... 27
2.6 Security, Crises and Military Applications ................................................................. 30
2.7 Recommendations ....................................................................................................... 33
Chapter 3. Introduction To Field Science ......................................................................... 34
3.1 Introduction................................................................................................................. 34
3.2 Remotely Deployed Sensors in Ecological Sciences .................................................. 34
3.3. Instrumenting the Environment: Pervasive Environmental In-situ Sensors .............. 38
3.4 Communicating in the Field ........................................................................................ 41
3.5 Other Field Applications: Weather and Agriculture ................................................... 43
Chapter 4. Sensors, Wireless, and Power Sources ............................................................ 47
4.1 Introduction................................................................................................................. 47
4.2 Overview of Sensor Technology................................................................................. 47
A. Cluster Sensors and Electronics ............................................................................... 51
B. Cluster Communications and Power ....................................................................... 54
C. Network Architectures.............................................................................................. 58
D. Design and Simulation of Wireless Sensor Networks......................................... 60
E. Emplacement and Deployment of Wireless Sensor Networks ................................ 61
F. Wireless Sensor Systems and Networks ................................................................... 62
4.3 Power........................................................................................................................... 68
4.4 Laser Wireless Communications for Data Transfer from Remote Sensors ................ 73
4.5 The Future -- A Mobile Internet Powered by a Planetary Computer.......................... 75
Chapter 5. Security, Privacy and Policy............................................................................ 81
5.1 Introduction................................................................................................................. 81
5.2 Overview of Security Issues........................................................................................ 81
5.3 Privacy and Ownership of Data .................................................................................. 84
5.4 Policy and the FCC ..................................................................................................... 85
5.5 Technology Transfer................................................................................................... 89
5.6 Communities ............................................................................................................... 90
The Military and National Security............................................................................... 90
Crisis Preparation, Mitigation, Management, and Response ........................................ 90
Distance Learning.......................................................................................................... 90
Health and Medicine ..................................................................................................... 90
Academic Research....................................................................................................... 90
page-iv
Technology Influenced and Enabled Genres of Art and Music.................................... 91
International and Cultural Exchange ............................................................................. 91
Remote Communication................................................................................................ 91
Commercial and Industrial............................................................................................ 91
5.7 Funding........................................................................................................................ 91
Chapter 6. Proposed Additional Workshops..................................................................... 94
Biographies of Speakers.................................................................................................... 96
Biographies of Participants ............................................................................................. 102
page-v
Chapter 1. Introduction
1.1. Overview Of Workshop
Several recent meetings have highlighted the growing interest of the biological,
ecological, environmental and other research communities in the development and
deployment of large distributed systems for information collection, aggregation,
analysis, simulation, predictive modeling and real-time analysis within their
respective disciplines. Likewise, recent events have caused concern and raised
interest in the need for similar systems in the context of homeland security. It seems
apparent that a structured discussion of the needs for these systems could create
opportunities for synergistic efficiencies between the various user communities. To
date, however, no such discussions have been held among these differing user
communities, nor among these communities and the groups providing and developing
technologies to support their needs.
It is with the possibility of identifying such potential synergies that the Multi-Sector
Crisis Management Consortium (MSCMC) is organizing a series of workshops to
discuss the design, development and deployment of such systems "From Sensor to
Supercomputer and Back: Systems for Information Collection, Aggregation,
Analysis, Assessment and Realtime Simulation." This workshop is the first in this
series. The purpose of this workshop was to summarize the current and projected
states of sensor, communication and power technologies for collection and
communication of information from distributed sensor systems with large numbers of
nodes, and to survey the present and expected needs for such systems in
environmental, facility, medical, crisis management, public safety and national
security arenas. A succinct summary of the current and projected technologies
mapped to current and projected user requirements was envisioned to result from this
workshop.
This workshop brought together experts in field sensors, signal processing, wireless
communications, and field power source technologies to structure solutions for the
common problems of collecting and communicating data from remote field locations
by cost-effective, scalable, integrated devices and networks. The workshop examined
the user needs and technologies involved in field deployable systems for information
gathering including sensors, wireless communications and power supplies, up to the
local aggregation point. While the integration of "locally" gathered information into
larger infrastructures for purposes of analysis is also required, such integration is
beyond the specific focus of this workshop.
This two-day workshop produced:
1. Understanding by participants of the current state of the art and availability of
technologies of sensors, power, integrated circuits and wireless communications,
and how they may be employed in our society.
page-1
2. Determination of current and future needs for better capabilities
3. Suggested courses of action to integrate available technologies into scalable,
reliable, cost-effective devices and networks of value to meet those needs,
including both the general scientific and research community, and government
agencies seeking better ways to protect the public from chemical and biological
incidents
4. Proposed design and production of devices not now available or suitable for the
above purposes
5. Recommendations for regulatory relief, especially for design, manufacture and
use of workshop-proposed wireless technologies for the above purposes
6. Recommendation for further work in these areas, and identification of potential
collaborations and partners.
The sensor systems considered in this workshop can be put in context by considering
our senses. We humans contain many sensors that transduce physical, chemical and
biological conditions into perceptions within our central nervous system.
Combinations of sensory organs, nerves for signal conduction and the brain, which
processes, stores and acts upon information from the sensors, respond to both internal
and external stimuli. We have many kinds of internal sensors. They provide
information on our positions, movements, blood chemistry, the degree of filling of
various organs and other factors. These internal sensors generally provide threshold
responses, and they are slow, that is, they have low bandwidth.
External Human
Senses
Temperature
Force and Vibration
Taste and Smell
Hearing
Vision
Human-Designed
Sensors
Many Stimuli
Max. Range for Receipt of
Stimuli
Contact
Contact
Contact
Kilometers
(Depends on Loudness)
Utility Over 10 Kilometers
Bandwidth of Response to
Changes
Less than 1 Hz
Less than 1 Hz
Less than 1 Hz
Greater than 1 Hz
About 30 Hz
Interplanetary
Over 1 GHz in Extreme Cases
(Communications Limited)
Table 1.1. Ranges and bandwidths for Human Senses and Designed Sensors
The external sensors respond to several well-known stimuli, as listed in Table 1.1.
The thermo-mechanical responses of the skin to temperature, forces, and vibrations
require contact with the factor being sensed, and they have low bandwidths in
response to variable stimuli. The same is true for the chemical senses of taste and
smell. The spectroscopic senses of sight and hearing have both relatively long
reaches and faster responses. Our external senses react to continuously varying
inputs over wide ranges, but do not provide quantitative measures of the various
stimuli. There are many conditions to which we are not intrinsically sensitive. The
page-2
Sensors Magazine Buyers Guide lists over 100 categories of designed sensors
[http://www.sensorsmag.com]. In most cases, these artificial sensors provide
quantitative information over their ranges of utility. As indicated in Table 1.1,
designed sensors can obtain information over long distances, depending on available
communications. They respond to rapidly changing inputs in many cases, and can
provide constant vigilance.
The emergence of designed sensors to compliment our natural senses has given us
important new tools. We can now respond to many more quantities and to harsh
conditions that would be unhealthy or lethal for humans. The availability of wired
connections (notably the Internet) and wireless communications (including
technologies like BlueTooth, 802.11b WiFi, cellular networks and satellite systems)
give our "new" senses much greater reach and bandwidth than our natural senses.
Basically, a flood of information from sensors around the world can come to us. It
must then be coupled into our brains primarily through our vision and hearing. The
limitations of our ability to assimilate information, and the fact that we must pay
attention to our surroundings even while focused on receipt of information from
sensors, have two important implications. First, information from sensors has to be
"down-sized" as much as possible as soon as possible during its flow from sensors to
humans. Consider our vision system as a megapixel imager that receives color and
brightness information from each pixel at rates exceeding 30 Hz. It must shunt
unimportant information, or else we could not assimilate the stream of input data. So,
our vision system has capabilities, such as motion and edge detection, which permit
us to focus on a small but important subset of the inputs to our eyes. In a similar
fashion, it must be possible to shed relatively useless information in cases of highbandwidth wired or wireless sensor networks. Second, because information can be
received from many sensors, there is the challenge of melding that information into
forms that can be consumed by humans without undo effort. This challenge is more
than "data fusion". It is really about "information fusion" in ways that match the
physiological, psychological and varied conditions and capabilities of a wide
spectrum of people. If sensor information goes to computers for storage or use in
control systems, it can be assimilated at rates exceeding what people can handle.
However, requirements for discarding useless information and for presenting it in a
proper fashion are also germane in those cases.
Since the employment of sensor
systems is about the acquisition of
information, it is useful to consider
the larger context indicated in Figure
1.1. Each of the five functions of
information technology shown has
many facets, and they evolve over
time. New aspects get added to each
major function as time passes. The
employment of sensors in large
numbers as a means of generating
PROCESSING
GENERATION
COMMUNICATION
By People, Computers
& Sensor Systems
UTILIZATION
& Diverse Systems
STORAGE
Figure 1.1. The major functions within
information technology and their
relationships. Sensors are rapidly growing in
importance as sources of information.
page-3
information is a comparatively recent phenomenon. Ten years ago, relatively little
information came from sensors. Ten years from now, a very significant fraction of
information will originate from sensors, especially because of the growing
commercial importance of sensor systems.
The current "Information Revolution" results from four micro-technologies:
electronics, magnetics, optics and mechanics. The connectivity between the major
aspects of information and micro-technologies is given in Table 1.2. Most of the
entries are evident. Optical computers remain under development and (still) promise
high speeds. Radio-frequency MEMS (MicroElectroMechanical Systems) are being
developed for both signal processing and communications. MEMS data storage
devices, which might offer terabit per square centimeter storage densities, are now in
development. There are some related and important aspects of information
technology not shown in Figure 1.1, including the reliability and security of computer
and communications systems. However, Figure 1.1 and Table 1.2 show the role of
micro-sensors based on different technologies, and the complex inter-relationships of
information and micro-scale technologies.
µ-Electronics
µ-Magnetics µ-Optics
µ-Mechanics
Computers
µ-Sensors
µ-Sensors
µ-Sensors
Computers
In Development RF MEMS
Wireless
Optical Fibers &RF MEMS
Communications
MEMS
Storage
Flash Memories Hard Drives Compact Discs In Development
Utilization
Control Systems Data Mining Displays
µ-Actuators
Table 1.2. Major IT Processes Correlated with Technologies in Sensor Systems.
IT/µ-Tech:
Generation
Processing
Communication
Millions of People
Given the great and growing
Wireless Subscribers
& Wireless Internet Subscribers
importance of the Internet, it
1 out of every 6 people on earth!
1000
is natural to question the
800
relative importance of wired
600
and wireless sensor
400
networks. The case for
200
wireless sensor networks has
two primary facets. The first
1998
1999
2000
2001
2002
2003
Figure
1.2.
The
growth
of
wireless
usage
is the growing availability of
worldwide.
wireless connectivity. We
are in the early stages of
what has been termed the "Wireless Revolution". BlueTooth, 802.11b WiFi, cellular
networks and satellite systems have already been noted above. The growing use of
cellular systems is especially striking. Figure 1.2 shows the increases in wireless
connectivity for both voice and data communications 1 . Sensors, and also hand-held
computers, have already been integrated with cell phones. It is only a matter of the
market before sensors, computers and wireless communications are all integrated into
1
The Industry Standard Magazine, p. 72, January 10-17 2000
page-4
units about the size of cell phones. The other driver for wireless sensor systems and
networks is the savings in costs for fixed systems, and the need for mobility in some
cases. Wires are expensive to install and protect or maintain. Fixed wireless systems
can be modified relatively easily. Issues of interference and reliability of wireless
systems, in general, and wireless sensor systems, in particular, are germane. Usually,
they depend on the specific case of interest. A recent poll by Sensors Magazine asked
at what stage respondents were in regarding implementing wireless networking1 . The
results were: 25.6% have no plans to implement a wireless system, 17.9% were
considering wireless connectivity, 20.5% had a plan for 1 to 2 years, 12.8% had a 712 month plan, 7.6% would make the move in a few months and 15.3% already have
a wireless network. It is likely that the experiences of people who implement
wireless sensor networks, especially in factories and other facilities, will influence
others to do similarly, as the systems are shown to be reliable, as well as costeffective. The number of companies that sell and service wireless sensor systems and
networks can be expected to grow significantly in the coming years.
This workshop is motivated partly by the likelihood that experience with the use of
wireless sensor systems and networks in one industry will be useful in other
industries. In general, each major industry has its own trade shows and publications,
so there is little routine communication between large industries. Despite the great
differences in the hardware deployed and the disparate use of the gleaned
information, there is a common base of components and concerns for wireless sensor
systems and networks. Such characteristics are reviewed during the workshop in
order to introduce the technologies that enable and limit wireless communication of
information from sensors.
1.2. The Program
The first day consisted of half-hour presentations by users presenting some key needs
and successes, and by technologists describing the state of the art and future
developments. The second day was dedicated to creating a report describing the state
of this field and the key steps that must be taken in order to make progress. All
available slides and videos may be found at
http://homeland.maine.edu/anywhere.htm.
1
Sensors Magazine, p. 60, July 2002
page-5
Day 1
PART I -- USER PERSPECTIVES
8:00 -- 8:15
Coffee, Registration
8:15 -- 8:20
Welcome -- Side Nadir, Executive Director, National Response
Center, and Acting Chair, Multi-Sector Crisis Management
Consortium
8:20 -- 8:30
Opening Remarks -- George Markowsky, Chair Departments of
Computer Science and Mathematics/Statistics, University of
Maine, and Secretary, Multi-Sector Crisis Management
Consortium
8:30 -- 9:00
Crisis Management -- Capt. Dennis Egan, Chief, Office of
Command Control and Preparedness US Coastguard, and Member,
Multi-Sector Crisis Management Consortium
9:00 -- 9:30
Biological & Ecological Sciences -- Tim Kratz, Associate
Director for Trout Lake Station Center for Limnology, University
of Wisconsin-Madison
9:30 -- 10:00
Environmental Monitoring & Public Safety -- Neil Gray,
Executive Director, International Bridge, Tunnel, Turnpike
Authority
10:00 -- 10:30
Break
10:30 -- 11:00
Facility Monitoring -- Richard Holm, Reactor Administrator,
Nuclear Reactor Laboratory, University of Illinois
11:00 -- 11:30
Communications -- David Hughes, Partner, Old Colorado City
Communications
11:30 -- 12:00
The Big Picture -- Larry Smarr, Professor, Computer Science
and Engineering Department, University of California San Diego,
and Director of the California Institute for Telecommunications
and Information Technology, Universities of California at San
Diego and Irvine
12:00 -- 1:30
Lunch
page-6
PART II -- TECHNOLOGY PERSPECTIVES
1:30 -- 2:00
Sensor Technologies and Applications -- David Nagel, Research
Professor, School of Engineering and Applied Science, The
George Washington University
2:00 -- 2:30
Sensors -- The Present -- Gregory Bonito, Research Assistant,
Long Term Ecological Research Network Office
2:30 -- 3:00
Communication Security -- Leslie Owens, Founder and CTO,
Vectara Innovations, LLC, Adjunct Assistant Professor,
Georgetown University and Technical Editor, Wireless Security
Perspectives
3:00 -- 3:30
Break
3:30 -- 4:00
Power Systems -- Robert Nowak, Program Manager, Advanced
Energy Technologies, DARPA
4:00 -- 4:30
Medical Systems -- Jim Wilson, WHO/NASA
Ebola/Marburg/VEE Remote Sensing Projects; GDIN Infectious
Diseases Working Group - Global Epidemic Intelligence and
Disease; Forecasting Systems (GEIFFS) and Pediatrician,
Georgetown University
4:30 -- 5:00
Regulatory Issues -- William Lane, Chief Technologist,
Wireless Communications Bureau, Federal Communications
Commission
5:00 -- 5:15
Closing Remarks
Day 2
PART III -- WORKING GROUPS
8:00 -- 8:15
8:15 -- 8:30
8:30 -- 12:00
12:00 -- 1:30
Coffee
Organization of Groups -- George Markowsky
The Technology Group, chaired by David Nagel and the User Group,
chaired by George Markowsky, worked through the material
presented on Day 1 and organized it so needs and possibilities were
highlighted.
Lunch
page-7
PART IV -- DRAFTING OF REPORT
1:30 -- 2:30
2:30 -3:00 -3:30 -4:00 -4:30 --
3:00
3:30
4:00
4:30
5:00
Presentations of Working Group Results and discussion –
David Nagel, George Markowsky, group
Outlining of Final Report -- George Markowsky
Break
Outlining of Final Report -- George Markowsky
Task Assignments -- George Markowsky
Planning for the Next Workshop & Close -- George Markowsky
1.3. Impact on Discipline
To the best of our knowledge, this is the first time that such a wide ranging group of
scholars and practitioners has been assembled to look at all aspects of sensors and
communication systems. This was driven by the unprecedented need to develop both
a more robust scientific network and by the needs for greater homeland security. This
workshop also attracted international interest, as evidenced by the participation of
Russian delegates. It is vitally important for us to better understand the capabilities
that others have developed in this area.
1.4. Dissemination Of Results
Information about the workshop, including the videos and slides of the presentations,
and this report can be found at http://homeland.maine.edu/anywhere.htm. They are
also mirrored at http://www.mscmc.org, which is the website of the Multi-Sector
Crisis Management Consortium. We also expect that, as a result of this workshop,
peer reviewed articles and research grant applications will be produced on the various
subjects covered. Many of the speakers and participants are extremely well known
and influential in their fields of expertise and we expect that the results of the
workshop will be disseminated by them in the course of their activities.
1.5. Pictures
Some pictures taken during the proceedings and of the ACCESS Center are available
at http://homeland.maine.edu and http://www.mscmc.org. In addition, PDF versions
of this report are also available at these websites.
1.6. Similar Workshops
The most recent, prior, workshop that was similar was held in San Diego, California
on October 29-31, 2001. It was organized by Greg Bonito who was one of the
speakers at this conference. Some details about this workshop are presented below.
page-8
SCALABLE INFORMATION NETWORKS FOR THE ENVIRONMENT (SINE)
Workshop Overview
The SINE workshop, hosted by the Partnership for Biodiversity
Informatics (PBI), will assemble research scientists, directors of
field stations and marine laboratories, as well as experts in
computational and information sciences to discuss the technical
requirements for building local, regional, and national-level
networks designed to deliver continuous, integrated high-quality
data in real or near real time. Each of these audiences will
contribute to and learn from the information exchange. Scientists
will share their experiences in expanding site-specific science to
broader spatial scales, and will discuss future information
infrastructure needs in light of new sensors (field and satellite)
and data collection capabilities.
Directors of field stations and marine laboratories will ground the
workshop in the present-day realities of existing infrastructure and
capabilities, and will contribute to a new vision of how field
stations and marine laboratories can expand to meet the needs for a
national capability for observing and understanding environmental
complexity. Computational and information scientists will present
state of the art developments in sensor technologies, networking,
information delivery, and knowledge generation.
Workshop presentations, discussion, and working group sessions will
focus on three topics:
1. Building distributed sensor networks: design and
implementation
issues.
2. Enabling technologies and user requirements for data and
information management and delivery.
3. Building scalable environmental information networks--data,
computers, and people.
As environmental research becomes more complex and
multidisciplinary, gains in our understanding of ecosystem
biocomplexity can be furthered through the application of
technologies that improve data management and delivery; enhance
modeling and prediction capabilities; and facilitate communication
among individuals, environmental sensors, computers, and databases.
This workshop will be the first attempt to envision a scalable
national environmental information infrastructure that meets the
needs of scientists working at local and broader scales, as well as
decision-makers and educators that may require information at
regional to national scales. Consequently, the discussions and
working group reports are anticipated to be of broad interest to
many disciplines. To meet this information need, several means will
be used to disseminate the workshop products to the broader
community.
page-9
Section 1.7. Acknowledgements
There are many people who made this workshop possible and we would like to thank
all of them.
First, we want to acknowledge the support of the NSF, which made a grant available
to fund the expenses of this workshop. In particular, Tom Greene, the program
director in charge of this grant was extremely supportive and helpful.
Second, the staff of ACCESS was of great help. Of special help were Tom Coffin,
Geoff Koslow, Jennifer Pillen, Kirstin Riesbeck, and Kia Ray.
Third, we thank the organizing committee of Dave Hughes, George Markowsky, Don
Mitchell, Dave Nagel and Janet Thot-Thompson.
Fourth, we thank all the speakers and participants for contributing their time and
expertise.
Fifth, special thanks go to David Nagel, Don Mitchell, Janet Thot-Thompson, and
Tom Williams, and John Porter for their assistance with the writing. David Nagel
wants to acknowledge B. Nickerson and W. J. Kaiser for some helpful conversations.
Fifth, we would like to thank George Brett and Dewayne Hendricks for some
suggestions in organizing the workshop.
This report mentions many products. Such references to products do not imply their
endorsements.
page-10
Section 1.8. Participants.
Alexander S. Akhmanov
Head of International Department, Ph.D.
Russian Academy of Sciences
Institute on Laser Information
Technologies
7-095-135-02-54
[email protected]
Neil A. Gray
Director of Government Affairs
IBTTA
202-659-0500
[email protected]
Rich Holm
Reactor Administrator
Greg Bonito
University of Illinois
Research Assistant
217-333-7755
Long Term Ecological Research Network [email protected]
Office
505-292-9331
Dave Hughes
[email protected]
Principal Investigator
Old Colorado City Communications
Michael R. Chritton
719-636-2040
Director of Security Programs
[email protected]
C2M Hill
303-771-0900
Brad E. Hutchens, P.E.
[email protected]
Environmental Engineer
U.S. Army Center for Health Promotion
Tom Coffin
and
Technical Coordinator
Preventive Medicine (USACHPPM)
ACCESS
410-436-6096/8162
703-248-0105
[email protected]
[email protected]
Anna Komarova
Capt. Dennis Egan, Ret.
Professor, Head,
US Coast Guard, now
Dept. of Foreign Languages for Geography
Director Homeland Security for
Lomonosov Moscow State University
Intermodal Transportation
095-932-88-35
System Planning Corp.
[email protected]
703-351-8289
[email protected]
Tim Kratz
Associate Director
Steve Goldstein
Trout Lake Station Center for Limnology
Senior Advisor of Information Technology University of Wisconsin
NSF
715-356-9494
703-292-4605
[email protected]
[email protected]
page-11
William D. Lane, Ph.D., PE
Chief Technologist Wireless
Telecommunications Bureau
Federal Communications Commission
202-418-0676
[email protected]
George Markowsky
Prof. of Computer Science, MSCMC
Secretary
University of Maine
207-581-3940
[email protected]
Dr. William K. Michener
Senior Research Scientist
The University of New Mexico
505-272-7831
[email protected]
Don Mitchell
Associate Director for Strategic
Collaborations
NSF/NCSA
703-2480123
[email protected]
Dick Morley
Consultant
R. Morley Incorporated
603-878-4365
[email protected]
David J. Nagel, Ph.D.
Research Professor MEMS and
Microsystems
George Washington University
202-994-5293
[email protected]
Bob Nowak
Program Manager
DARPA
703-696-7491
[email protected]
Leslie Owens
President
Vectara Innovations
703-980-3377
[email protected]
Jennifer Pillen
Administrative Assistant
ACCESS
703-248-0104
[email protected]
John Porter
N/A
University of Virginia
434-924-8999
[email protected]
Syed Qadir
Executive Director (MSCMC Acting
Chair)
National Response Center
202-267-6352
[email protected]
Kirstin Riesbeck
Program Manager
ACCESS, Shodor Foundation
703-248-0120
[email protected]
Joann P. Roskoski
Deputy Division Director
Division of Environmental Biology
National Science Foundation
703-292-8480
[email protected]
Frank Sellers
President
NAFLET
703-931-2862
[email protected]
page-12
Bill Taylor
President
Virtual West Point, Inc.
443-742-4518
[email protected]
Shukri Wakid
Senior Scientist
Compaq
301-237-9276
[email protected]
Janet Thot-Thompson
Associate Director
MSCMC Acting Executive Director
ACCESS
703-248-0102
[email protected]
Tom Williams
Principal
Air Networking
804-672-7386
[email protected]
Frank Vernon
Research Geophysicist
University of California, San Diego
858-534-5537
[email protected]
James Wilson, V, MD
GDIN Infectious Diseases Working Group
GDIN
202-270-2527
[email protected]
Brief biographies of the speakers and participants are in the Appendices.
page-13
Chapter 2. Introduction To Homeland Security
2.1 Introduction
The talks by Syed Qadir, George Markowsky, Dennis Egan, Rich Holm and Jim Wilson laid out
most of the issues dealing with homeland security. Syed Qadir, Executive Director of the National
Response Center located at the U.S. Coast Guard Command and Control Center at Buzzards Point,
Washington D.C., opened the workshop with an introduction. Syed Qadir is also the Acting Chair
of the Multi-Sector Crisis Management (MSCMC) Consortium, which was founded in July 2000.
He noted that this was just the first of a series of workshops hosted by the MSCMC on the synergy
of sensors and supercomputers. We will now quote from the talk. All of these talks with supporting
slides can be viewed at http://homeland.cs.umaine.edu/anywhere.htm.
The MSCMC was formed nearly two years ago, when several of us in the field of crisis
management recognized a need to get people together. It was clear that amazing
technologies were available to aid crisis response, but those technologies, because they
were spread out and fragmented among the private, government, academic, international,
and nonprofit organizations, were not being developed in a coordinated manner. Since its
formation, the Consortium has attempted to break down communication barriers between
these various sectors and bring them together to make sure everyone is working together
and using the best tools in the best ways to efficiently and safely manage and respond to
crises.
With these goals and objectives in mind, we turn first to sensors and supercomputers, as
topics of great importance and a telling example for the need for collaboration and
sharing. Data is of little use without tools to aggregate, analyze, communicate, and apply
the data. Supercomputers and communication technology, on the other hand, capable of
organizing and distributing vast amounts of information, are also of little use without
accurate and timely data, thus inspiring the title for this first workshop. This workshop has
brought together representatives of multiple sectors and diverse backgrounds. Sitting
among you, of course, are top people in the fields of sensors and supercomputers, but there
are also experts in other sectors and areas including lasers, languages, disinfection,
disasters, governing states, global communications, chemical defense, collaboration,
biogeochemistry, broadband data devices, wireless security, and water ecology.
With this brief introduction, we hope that you, with your varied perspectives, your
combined creativity, and wide-ranging knowledge will illuminate new ideas and
applications, identify and propose solutions to current challenges, and envision the future
of data collection and analysis technology-that ultimately you will expand the field of
sensors and supercomputers for global peace and prosperity and democracy.
page-14
2.2. The Big Picture
George Markowsky gave some background on major terrorist events and the creation of the
MSCMC. The Multi-Sector Crisis Management Consortium (MSCMC) was created in July 2000.
It was chartered as a Maine non-profit organization, with its headquarters here at the ACCESS
Center. MSCMC has the following goals:
1.Identify the needs of the crisis response communities.
2.Support advanced information technology research for crisis management, emergency response
and related communities.
3.Bring multiple sectors together.
4.Provide neutral convening forum.
5.Provide a portal for user communities to the R & D communities.
A major event that provided impetus for the organization of the MSCMC was the March 20, 1995
Tokyo nerve gas attack by Aum Shinrikyo cult. One of the most frightening aspects of this attack
was its sophistication. Multiple locations were attacked simultaneously, and the cultists also tried to
interfere with the first responders. Two books that provide much more information about this
attack are Underground: The Tokyo Gas Attack and the Japanese Psyche by Haruki Murakami
and Aum Shinrikyo-Japan's Unholy Sect by Rei Kimura.
In the article Combating Terrorism at Home by Douglas J. Gillert in the Journal of Civil Defense,
Winter 1998, the author describes crisis responders by the phrase:
“The brave, the few, the underequipped, the undertrained
He notes that combating domestic terrorism falls on the shoulders of emergency services and other
municipal agencies, but that these agencies are not equipped or trained to handle this
responsibility.
The focus of the MSCMC is not exclusively on terrorism, but on crisis management in general. In
our increasingly more complicated and interconnected world, very serious accidents can happen. A
particularly chilling example was the December 3, 1984 accident at the Union Carbide plant in
Bhopal India that killed nearly 4,000 people.
In addition to terrorist acts and accidents, we live in a world in which natural disasters can happen
in most locations. Natural disasters have great potential for creating disasters. A particularly
frightening form of natural disaster is an epidemic. Epidemics have broken out from time to time,
and claimed many lives. The ever-changing genetic structures of microbes and viruses hold open
the possibility of future large-scale epidemics.
The book Hot Zone by Richard Preston describes a 1989 episode in Reston, Virginia involving the
ebola virus . Fortunately, the strain involved in this incident only infected chimpanzees. Otherwise,
there would have been an outbreak of ebola virus in the suburbs of Washington DC. As travel
increases and human density grows, the chances of a truly devastating epidemic arising because of
accidental contact between humans and microorganisms increase.
page-15
Serious incidents might involve multiple factors. The book Normal Accidents by Charles Perrow,
discusses the concept of a normal accident, which is an undesirable event that is truly nobody’s
fault. Normal accidents are a consequence of unexpected and unintended interactions. An amusing
instance discussed by Perrow was an accident involving a salt mining company and an oil drilling
company whose independent actions a lake to disappear.
The book , The Monkey Wrench Gang by Edward Abbey has inspired a small number of
"ecoterrorists" who have acted against various experiments. Even in Maine attacks have taken place
against fields of new plant species. It is conceivable that an attack against a laboratory might lead to
an accidental release of some agent that could have devastating consequences.
The MSCMC is ready to play its part in making the world safer for humanity. Its strong partnership
with the ACCESS Center here in Arlington Virginia, shows some of the technology that can be
effectively harnessed in the fight against disaster. In May 2001, the ACCESS Center successfully
served as the home base for the TOPOFF (Top Officials) Exercises, which simulated setting up an
alternative communications facility in the event of three major incidents (biological, chemical,
nuclear) occurring simultaneously in widely separated areas of the country. Among other projects,
the ACCESS Center and its affiliates are investigating how to make the ACCESS technology
portable. This project is called: AccessGrid to Go. Additional information about the MSCMC and
its events is available at http://www.mscmc.org.
2.3 The Threat to Homeland Security
Captain Dennis Egan, our second speaker, began with the observation that the perception of the
world has changed remarkably in the last several months. He went on with the following
observations.
The world seems to be a much more dangerous place, but it might be that the world has not
changed, and that only our perceptions of the world have changed. Rather than a conventional
belligerent nation, we are confronting a more ancient foe.
The nature of the beast is global and supra national in scope (outside the construct of the nation
state) with a dangerous unifying doctrine of hate, intolerance and violence. Not since the days of
the anarchists has the world experienced such a threat, and the important difference today is these
terrorists have access to weapons of mass destruction. Starting with the Persian Gulf War of 19901991, the United States ushered in a new type of war with a focus on destroying critical
infrastructure of the enemy using both conventional and somewhat revolutionary means. This was a
war that ushered in the use of stealth fighters and bombers, cruise missiles, and a satellite controlled
battlefield with a particular emphasis upon destroying the enemy's command, control and
communication systems, including the ability to control and distribute electricity and fuel and
water. This was the introduction of "nodal" warfare, destroying the enemy's ability to use critical
infrastructure.
page-16
During the decade that followed, we started to realize that our own nation, by becoming
increasingly complex, tightly coupled and heavily dependent upon computerized control and
communications, was becoming more vulnerable to cyber attack and cyber disruption.
Considerable emphasis was placed upon introducing continuity of operation plans that ensured
redundant and appropriately isolated systems. This effort peaked during Y2K which, being a nonevent, convinced most of the public that our vulnerabilities were not as bad as they might have
seemed at first.
As almost an afterthought, the press reported on a foiled terrorist plot to destroy the Seattle Space
Needle by Algerian terrorists linked to the Osama Bin Laden organization. Few understood the
nature of this linkage, however some understood that there was some indirect linkage to terrorists
who had blown up two US Embassies in Africa, destroyed American soldiers' barracks in Saudi
Arabia, and carried out an abortive attempt to destroy the World Trade Towers.
Following the attacks on the World Trade Towers and the Pentagon on the 11th of September, we
realized that these terrorists did not have to bring their weapons to the party, they used an inside-out
strategy which turned the risks of our own transportation systems against us. One should never
forget that the most effective weapon against the might of the US Navy, was the suicidal-manned
aerial bomb, the kamikaze. Indeed, much of our critical infrastructure is subjected daily to risks
from the hazardous cargos of our nation's transportation systems. The public accepts the risks
because the carriers have employed very safe practices in transport operations that usually prevent
accidents from happening. This risk-tolerance on the part of government and the public has resulted
in some adverse linkages of hazards that in an unfortunate sequence can change the occurrence of a
number of unrelated accidents into a tightly coupled chain reaction, with disastrous consequences.
For instance, some ports employ the practice of segregating ships carrying explosive cargoes into a
common anchorage, sometimes with little recognition of adjacent shore handling facilities that have
large quantities of toxic materials stored on site. The fact of the matter is that if one ship is
exploded by a terrorist, perhaps by the crash of an airplane, several other ships and toxics ashore
can be ignited with dire consequences to the coastal population.
The USCG Commandant, in a recent speech to the International Maritime Organization (IMO),
emphasized the need for security plans that address vulnerabilities of individual facilities, offshore
terminals and ships. Had the IMO been a multi-modal, all-hazard organization, he could have gone
further in recognizing the close coupling among hazards, including transportation systems that
provide incentives for exploitation by terrorists, because they amplify the consequences of a
terrorist attack into a much larger complex disaster. Despite the creation of many individual
security plans that emphasize preparedness and rapid reaction or response, unless they are
correlated into a larger regional or area infrastructure security plan, that recognizes the potential
linkages among hazards, terrorists will maintain an advantage of choosing the weakest or most
vulnerable link with greatest systematic consequences.
Terrorists use our own transportation systems against us exploiting our freedom of movement that
creates significant economic advantages of free markets in the "Developed world". Such a concept
greatly influenced the Asian Pacific Economic Conference "Economic Leader's Statement On
Counter-Terrorism" 21 October 2000 in Shanghai, China. APEC emphasized the need to coordinate
page-17
plans and preparedness to enhance airport, aircraft, and port security. It recognized key linkages
between transportation system disruptions and temporary supply disruptions and longer-term
challenges facing the region's energy supply. APEC recognized the need for a coordinated effort in
critical sector protection including telecommunications, transportation, health and energy. It called
for the enhancement of international communications networks and accelerated sharing of
information among Customs and Border police organizations. It called for new response initiatives
to strengthen capacity building and economic and technical cooperation to help member economies
to put in place and enforce effective counter-terrorism measures. Finally, APEC called for greater
cooperation to limit the fallout or secondary effects of the attacks, and for the recognition of the
importance of recovering economic confidence in the region through growth-building coordinated
policies as well as providing a secure environment for recovery of trade, investment, travel and
tourism.
Note the much broader considerations of the APEC Economic Leaders' Statement. They called for
much more than a simple initiative on improving reactive forces or more completely sharing police
intelligence about terrorists. They called for an integrative approach with improved coordination in
the four pillars of disaster management that are Prevention, Preparedness, Response and Recovery.
The economic leaders used the language "coordination". What is really needed is more
collaboration than coordination. We find ourselves in much better shape to coordinate safety than
security. Security is quite different than safety. It is highly sensitive because it requires a great deal
of trust to divulge one's vulnerabilities even to a partner. Consequently, coordinating one common,
unified security approach is uncommon. Everyone can subscribe to safety, and work towards both
standards and international protocols. Because resources are limited, security requires that certain
national decisions be made about protecting some things more than others. Because a terrorist will
probe a potential target and select targets of lower resistance that still meet the terrorist's objectives,
it is unwise for a nation to openly set patterns of security protection and procedures "in concrete".
Security demands a degree of covertness while safety is an open discussion. Security requires
collaboration over secure communications while safety can be openly communicated and
coordinated.
Collaboration is essential to enhance and share security technology that provides such initiatives as
integrated surveillance and detection systems, alert and warning systems, secure communication
systems -- both wide bandwidth and wireless -- and decision support capabilities such as hazard
simulation and consequence modeling, pattern recognition, and rapid identity authentication.
Collaborating to establish common approaches to analyze vulnerabilities and understand the
relationships among various linked hazards, can lead to very productive initiatives to mitigate the
consequences of terrorism when it occurs. Often this kind of analysis can establish a potential
causal chain of events and heighten the awareness and pattern recognition should a terrorist
organization conduct a "dry-run" or the initial stages of an attack. This kind of analysis can also
suggest productive strategies to break the causal chain by reducing or eliminating coupling of
hazards. Such mitigating approaches are also beneficial in the recovery phase to ensure that
communities and their economies become more capable of sustaining and recovering from terrorist
attacks.
The Multi-Sector Crisis Response Consortium (MSCRC) is positioned in an extraordinary manner
to enhance the collaboration needed for combating and mitigating terrorism from international
page-18
organizations. It provides a wide bandwidth, common meeting ground to discuss and evaluate key
technologies such as Geographic Information Systems, which effectively communicate the
distribution of critical infrastructure and sensitive areas of the economy, modeling and simulation
of hazards to establish coupling and consequence footprints and development of econometric
models to establish the real costs both immediate from the terrorist attacks as well as that selfimposed reactive cost of initiating alternative counter-terrorism strategies. We have the challenges
of moving such discussions increasingly behind encrypted means. I enjoin this organization with
this imperative. During the recent APEC Transportation Security Experts Working Group session
in Singapore, I submitted a potential project proposal that drew heavily upon the capabilities of
enhanced collaboration offered by the MSCMC. This proposal would lead to both team-to-team
direct collaboration as well as video-teleconferenced collaboration in development of a toolbox of
mutually accessible models and simulations of hazards (transportation hazards with potential
danger to critical infrastructure) that might be unleashed by terrorists. The nations of Russia, China
(including Hong Kong), Singapore, Canada, Australia and New Zealand, were not in favor of
divulging their own appraisals of their nations' vulnerabilities. However, they were strongly in
favor of developing common techniques of analysis of the consequences of terrorist attack and a
community's capability to survive such an attack. Using a notional nation, the proposal would have
a multi-national team set up the GIS clearly demonstrating typical vulnerabilities and coupling of
hazards and critical infrastructure. It would then design a tabletop exercise that would engage the
power of collaborating over wide-bandwidth video teleconferencing and streaming video.
Participants would demonstrate using the toolbox as decision support for response to simulated
terrorist attacks as well as a means for improved prevention, preparedness, and community
recovery. The tabletop exercise could feature a "red" terrorist cell as well as defenders or terrorist
attacks could be pre-scripted. Discussion would draw on the predictions of the consequence
assessment tools used as decision support as participants would submit their proposed response and
mitigation strategies. A basic and relatively simple econometric tool could be used as a decision
support aid by participants to demonstrate the necessity of seeking a balance between security
imperatives and the economic benefits derived from free flow of trade. This conceptual proposal
was selected as one of the top four by the APEC Security Experts Group from among 12 proposals
submitted by various nations. It is being redefined this week as a proposal for a Phase 1 feasibility
study and is to be considered by the APEC Transportation Security Committee during its meeting
in Manila on 4 March 2002. Russia, China, Singapore, New Zealand and the US strongly supported
it at the Working Group level.
To conclude, the current focus on international counter-terrorism really demands a more
comprehensive approach to the four pillars of disaster management - Prevention, Preparedness,
Response and Recovery. While traditional approaches are used to coordinate international safety
initiatives, approaches towards strengthening our collective counter-terrorism capabilities require a
different sort of collaboration, more akin with traditional disaster management considerations, but
with some important differences in secure communications and respect for national sensitivities.
Key technologies are those that can enhance collaboration and visual communication in a secure
environment. New insight may be gained through creating a notional geographic information
system model and conducting team-based, collaborative consequence assessment analysis. This
would engage a common set of hazard simulation tools and an econometric tool for more effective
decision support. Table top exercises and follow-on discussions would emphasize the need for
page-19
strategies that enhance the ability for communities to both survive and recover economically from
terrorist attacks.
2.4 Transportation, Buildings and Homes
In his talk, Neil Gray, Executive Director of the International, Bridge, Tunnel and Turnpike
Authority, introduced himself and his colleagues as the "toll guys" - the people who worry about
places that have tolls. This includes many roads and bridges worldwide. He asked his Association
members what did they want to know? He found out that they are very eager to use sensors to
monitor axles, determine dimensions and read tags. They must have accurate information.
A big issue is vehicle framing, i.e., the identification of distinct vehicles. It is not always easy to
tell the difference between a trailer and a tailgater. He discussed various types of sensors that are
used to monitor roadways and talked about the problem of vehicle framing. As a vehicle travels
through a toll lane various sensors count axles, determine dimensions, read tags, etc. Vehicle
framing refers to insuring that all this sensor data is attributed to the correct vehicle. Common
vehicle separators used in toll lanes and their faults regarding vehicle separation were discussed. He
then talked about loop detectors, which have been around a while, and are very good at detecting
vehicles but not very good at separating them. Cars bunched close together, as happens in a toll
lane, cause the loop to remain on between vehicles. Vehicles pulling trailers (with very little ferrous
mass for the loop to detect) may have a dropout between the vehicle and trailer therefore spoiling
the vehicle separation. Hi-tech motorcycles made mostly of aluminum can cause the loop to miss
them completely. Very high vehicles such as tractors pulling trailers may also cause the loop to
"drop out" while the vehicle is still in the lane.
Sonic sensors, which are normally mounted looking down on traffic. These sensors often miss the
space between vehicles, as the sound sent from the sensor covers such a wide area. Much like the
loop detectors, they are good for detecting the presence of vehicles but not vehicle separation.
Through beam sensors do a much better job at separation since the light beam is emitted from one
side of the lane and received on the other side. This creates a pencil thin beam which when broken
detects vehicles. The problem comes with different height vehicles not triggering the beam. If the
beam is too low, it will not detect high trailers (except for the axles). If it is too high it may miss
lower vehicles such as a flatbed truck with no load on it. These sensors are susceptible to inclement
weather such as snow or rain. In addition, they degrade in performance as they collect dirt and dust.
Light curtains are essentially a vertical array of through-beam sensors. These have all the
advantages of the through-beam sensor and eliminate most of the disadvantages. The trouble with
light curtains is the price. Since they are based on through-beam sensors, the light curtains are also
susceptible to inclement weather, such as snow or rain. In addition, they degrade in performance as
they collect dirt and dust. Since a light curtain starts very close to the ground, it tends to be affected
by snow accumulation and snow plowing.
Laser sensors bounce a beam of light off a vehicle and read the returning beam. This works well in
a controlled environment such as a production line where the target will always be the same shape
and size unlike in a toll lane. These sensors are also susceptible to inclement weather such as snow
page-20
or rain. In addition, they degrade in performance as they collect dirt and dust. In addition, a laser
sensor has high precision moving parts and optics, which will eventually need replacement,
calibration, and other attention.
What people monitoring roadways need is a "dream" sensor that would have the following
qualities:
•
•
•
•
•
•
•
•
•
•
It could accurately separate vehicles, which are as close as 12 inches apart (closer would be
better).
It could recognize a very small target such as the trailer hitch of a vehicle pulling a trailer.
It would be impervious to weather and temperature extremes.
It would have a simple interface, such as a switch output indicating vehicle presence.
It would be simple to mount.
It would have low maintenance.
It would have extensive error reporting.
It would be extremely reliable through design or cost effective redundancy.
It would have a reasonable cost.
It would easily respond to active transponders, which are used as easy tags.
In addition to the need for improved sensors, the members of the authority are interested in issues
of related to shutdowns. Currently, they monitor temperature, weather, turn ramps, and the saline
content of water on the road. They are very interested in automated remote signage.
They need to monitor vehicle flow, which would lead to the creation of large temporary databases.
The goal is to purge records relatively rapidly and not to produce a national ID system. There is no
overarching entity in most areas that would be able to construct a national database. The different
members of the Association collect different types of data. There are no plans to integrate data
collection and analysis.
Another issue is monitoring air quality and determining what toxins are in the air. One unresolved
question is whether it is better not to know about toxins in your air. If you know about them, are
you liable? What can you do about it?
It would be great to have a sampling system for vehicles so you could know what is in a threat area.
Such a system could be linked to INS and Customs. There was an incident in which a truck burned
under a major overpass, which created traffic problems on the East coast for several days. It would
also be wonderful to have some system for counting the number of people in a vehicle, and to
determine some attributes of vehicles such as their weight, and get some idea of whether they are
carrying dangerous materials.
The Association's experience suggests that it is a long, expensive process to determine what
technologies work. Members always want something pervasive and inexpensive. They are very
willing to entertain new things, if they can justify the cost. The members do have significant sums
available, but their interest falls off if data does not come back in useful ways.
page-21
There are many issues that need consideration at this time. Some include being able to predict
problems on roads, bridges and in tunnels, and being able to ameliorate them. These include trying
to predict icing conditions, and terrorist attacks. Predicting attacks is only part of the problem -- we
must still determine what to do about them. As we develop technology, we must also be concerned
about privacy and costs. It was noted that perhaps the Orange County Expressway is one of the
most advanced systems in the world. However, they are blessed with generally excellent weather,
so their system might not operate as well in other environments.
Rich Holm, the reactor administrator for the University of Illinois Nuclear Reactor Laboratory,
talked about the trials and tribulations that would be associated with installing sensor technologies
into a 60 story commercial building. The primary question motivating him is what can be done to
make a building the safest commercial building ever built? This question suggests others, such as:
What can make it safe for occupants and for disaster responders? What are the threats?
These questions lead to the problem of deciding on the types of sensors to include (structural, life
safety, security, CBN, etc.) and figuring out how to power them. This is complicated by the fact
that some sensors will be “embedded” and inaccessible after construction. Associated with these
issues are questions of how to assimilate and present the information collected. Many of the sensors
that are required for monitoring transportation infrastructure need to be powered by batteries. Thus,
we would like to have batteries that would last at least 10 years before needing to be replaced.
Some sensors could be hard-wired to power, but this adds to wiring complexity and costs, which
already add 10% to the building cost. Undoubtedly, some concessions will have to be made. The
power source will most likely not be nuclear.
Rich Holm discussed the various types of sensors that might be used in a building. These include
structural sensors that measure skin pressure, temperature, displacement, and strain. Another class
is life safety sensors such as smoke detectors, thermal detectors, chemical and biological sensors,
and radiation measuring devices. A closely related class includes security sensors such as infrared
sensors, motion detectors, individual occupant sensors, and even facial recognition sensors.
We must realize that any building that we build at this time would literally be a living laboratory. It
would be essential to move data from the building to laboratories where it could be analyzed
further. It would also help us realize what items we have overlooked.
An important issue is that of survivability. Given that our vigilance will fail sometimes, we need to
know how to keep the system operating when something happens to it. This is still a wide open
question, which needs a lot of work.
We are learning a lot from Navy ships. If you rotate an aircraft carrier 90 degrees, it will look a lot
like a tall building. The next generation of aircraft carriers might have something like 0.25 million
sensors! Fortunately, the Navy has enough money to carry out the development of such complex
systems. Also, the Navy is very concerned with survivability.
We are also learning a lot from the nuclear industry. Partners such as architects have good archival
design information. The American Institute of Architects (www.aia.org) is a good source to tap
into and they have a sensor group. There are also related groups at Sandia National Laboratory
page-22
(www.sandia.gov). Fire departments and other first responders are also good sources of information
and ideas.
We conclude this section with some material supplied by David Nagel, whose talk is described in
more detail in Chapter 3.
Placing wireless sensors in protected environments
avoids many of the problems with the employment
of such sensors outdoors. However, some of the
systems associated with factories, facilities,
buildings and homes involve a mix of outdoor and
indoor sensors. Unlike the situation outdoors, Figure 2.1. Photograph of an uncooled infrared
from Raytheon with a micro-machined array of
power is generally available within or near camera
120 by 160 pixels, and a picture taken with the system.
structures. If electrical cables are run to sensor
nodes, then it is possible to employ the same cables to transmit data at frequencies well above those
found due to power transmission. Nevertheless, the use of wireless systems for delivering the
sensed information remains attractive in many cases. The monitoring of the condition and
performance of many systems associated with structures is the focus of this section.
The high costs of providing, using and disposing of energy can be ameliorated by the use of energy
monitors. The leakage of heat from buildings can be sensed with micro-machined infrared imagers,
such as the one in Figure 2.1, which view the building from its exterior and convey information to
the user in the building (or elsewhere) by wireless means. The presence of unduly high or low
temperatures within a structure can be determined using relatively simple sensors. The loss of
conditioned air to the exterior can also be determined with point sensors.
Electrical power distribution systems can be monitored for their state of operation as well as for
their energy efficiency. The same is true for fluid distributions systems. These include the HVAC
systems that move air, and the gas distribution systems in some structures, as well as the liquid
systems for water and waste. In general, computers control HVAC systems in response to
programmed instructions, which commonly vary with the time of day and day of the week.
Automatic sensing and control of gas and water systems in emergencies is provided in some cases.
Communication and information systems also offer
opportunities for monitoring. Their integrity and condition is
increasingly critical to the operation of factories, facilities,
buildings and even homes.
Sensing the integrity of such
systems is less complicated than sensing the conditions of
electrical and fluid systems because parameters such as power
and flow rates are not of interest.
Monitoring the condition and operation of machinery has
received great attention is recent years 1 . The goal is to reduce
maintenance costs by doing only what is needed when it is
needed. Parameters of interest include operating temperatures,
1
Figure 2.2. A wireless torque
sensor made by Binsfeld
Engineering
Society for Machinery Failure Prevention Technology, http://www.mfpt.org/
page-23
which can be sensed with infrared imagers and point detectors. Vibrations in many machines are
sensitive indicators of the alignment of the systems and its performance. The torque within drive
shafts is another factor of interest. Its determination is an excellent application of battery-powered
wireless sensor systems, which avoid the need for contact with the shaft to receive power or
transmit information. Figure 2.2 shows an example of a commercial system for measuring the
strain and, hence, the torque in a shaft during operation1 . Other companies also make wireless
strain gauges 2 and general multi-sensor systems 3 .
The control of steps within industrial processes that make materials, devices and more complex
products is a major area of applications for sensors within control loops. Wireless communication
of the information from the sensors to the computer, which controls some part of the process, is
being used increasingly. Similar systems can also be employed to assay the quality of the materials
or other products that result from a process. There are two primary reasons why wireless systems
are not yet more heavily used in industrial control loops. The relatively recent availability of
wireless sensor systems means that there are still many installed legacy control systems that are
wired. The secondary reason is the relative robustness of wired systems, that is, their immunity of
electromagnetic interference, especially for systems that have optical fiber links. It is likely that,
several decades from now, a significant fraction of industrial networks for transmission of sensor
information will still be wired. The situation may prove analogous to the penetration of small
computers into factories and offices. For all of their advantages, medium and large computers still
play a role in industries.
The habitability of buildings depends on the conditions of the air and the absence of deleterious
chemicals and organisms. Hence, many of the parameters of interest for outdoor weather are also
germane to indoor conditions. The same is also true for sensing the chemistry and biology of the
air, water and particles in structures. Probably, major buildings will have sophisticated chemical
and biological analyzers installed for both health and safety reasons, when such systems become
affordable and reliable. Here, also, there are historical precedents. Air conditioning was once
unavailable, but now it is routinely a part of buildings.
In the past decade, the concept of "smart" spaces and homes has arisen in parallel with the
emergence of "smart" sensors and networks. 4 The variety of things that are done in homes
significantly exceeds the functions in factories, facilities and many other buildings, notably offices.
Safety and security are a concern for the occupants of all structures, of course, and communications
is a common function within all structures. In addition, eating, resting, many kinds of cleaning and
maintenance, entertainment and other functions are all done within homes. Currently, sensors are
elements of the HVAC system control loop, but are not in many other home systems. Sensors are
1
D. Hebert, "Wireless Solutions Emerge", Control and Design, p. 45, May 2001 and http://www.binsfeld.com
Model 9300 Rotary Torque Measuring System, http://www.wirelessdatacorp.com/ wdc/index.html, and Strain Link ,
http://www.microstrain.com/slink.html
3
HiDRA: Highly Deployable Remote Access, http://www.rsc.rockwell.com/html/ information.html#Wireless Systems
& Application, and Wireless Web Sensor Networks, http://www.microstrain.com/WWSN.html.
4
B. R. Badrinath and M. Srivastava, "Smart Spaces and Environments", http://www.comsoc.org/livepubs/pci/
public/2000/oct/current.html, J. M. Sanders, "Sensing the Subtleties of Everyday Life", http://www.gtri.gatech.edu/rhwin00/main.html, Home of the 21st Century, http://www.seas.gwu.edu/~h21c/index2.htm and G. Aboud, "The
Broadband House: Related Links", http://www.emory.edu/college/scienceandsociety/scienceinyourlife/
broadbandlinks.htm
2
page-24
now found in appliances that involve high or low temperatures, including ovens, toasters, clothes
dryers, refrigerators and freezers, as well as water systems such as clothes and dish washers.
Monitoring other consumer products and even food is in prospect. Some homes also have sensors
for security systems. It is expected that sensors, central information systems and voice recognition
technology will be commonly found in homes in the future. These capabilities will enable a better
awareness of the activities and even the health of the occupants of the building. The information
can be used both within the home and beyond it. The current situation regarding "smart" homes
can be likened to another major consumer product, the automobile. Initially, advanced technology,
such as automatic braking systems and night vision devices, were found only on the most expensive
models. As the sales volume of smart home systems increases, prices should drop to where they
are more widely affordable.
In addition to monitoring the contents and occupants of structures, there is great interest in sensor
systems to assess the structures themselves. Responses to earthquakes and high winds, and the
possibility of flooding or gas leaks due to pipe breaks, are primary drivers for this interest. Strain
sensors are being installed in many new buildings to better understand what is needed for safe
design and construction of structures, that is, for improved determination of safety factors.
Current terrestrial, marine and air, and future space transportation systems are all candidates for
monitoring using wireless sensor systems and networks. There is great interest in "intelligent"
highway systems in Europe, Japan and the U. S1 . Motivations for making more effective highway
systems include increased safety, saving travel time and reducing pollution. Knowledge of the
current traffic load and speed over wide geographical regions is needed. Imagers can provide such
data if they are combined with sufficiently capable computers that have adequate analysis software.
Point sensors can also be employed to provide traffic information. This is already being done at
4000 locations on the German autobahn2 . The data obtained from sensors mounted on overpasses
is relayed using the cellular phone system. Monitoring conditions of and within train cars can be
done with wireless systems. An example graphic from such a system is given in Figure 2.3 3 . Air
traffic control is a very different challenge
than following cars, trucks or trains.
Currently, radars on the ground and in
larger planes are employed to sense the
presence of planes in a region. The use of
infrared imagers and acoustic sensors in
wireless systems on and near airports
might impact safety, traffic control and
noise abatement.
Figure 2.3. Map of the U. S. showing the
Automobiles and trucks have a growing location of trains equipped with sensors
number of micro-sensors in order to that report via satellite to a General
improve performance and safety. Most of Electric Service Center.
these are in wired systems within the
vehicles. The use of wireless systems to monitor tire pressure is expected. Under-inflated or over1
2
3
C. Schmidt, "The Road Ahead", MIT Technology Review, pp.73-77, July/August 2001
DDG Gesellschaft für Verkehrsdaten mbH, http://www.ddg.de
R. Pool, "If It Ain't Broke, Fix It", MIT Technology Review, pp.64-69, September 2001
page-25
inflated tires, especially on trucks, significantly increase operating costs. Wireless monitoring of
the conditions within semi-trailers is also in prospect, especially in refrigerated vehicles. Using the
Global Positioning System and wireless reporting is already done to follow the location of trucks
and train cars. Wireless sensor systems can also report on the conditions within shipping crates and
containers.
Wireless systems of sensors enable the operation of robots of many kinds. Robotic vehicles used
by police and others on land now include imagers for inspection of potentially dangerous objects.
They could also be equipped with additional sensors to determine the nature of unknown
substances, including nuclear materials, explosives, chemicals and biological materials. Unmanned
air vehicles are often used as platforms for imagers and other sensors. Unmanned underwater
vehicles can also be used as sensor platforms, from which the wireless link uses acoustic rather
than radio-frequency information communication techniques. The floor of the earth's oceans is
poorly known, and there is growing interest in robotic exploration for both scientific and practical
reasons. Project NEPTUNE envisions equipping much of three small tectonic plates off the coast of
Washington State with, first, wired (fiber optic) and, later, wireless (acoustic) sensors 1 . The
proposed global ocean-mapping project (GOMaP) would put to sea large numbers of underwater
vehicles to map the entire sea floor with side-scan sonar and other sensors 2 . The recorded data
would be sent through satellite systems when the vehicles come to the surface to recharge their
batteries.
Robotic vehicles will also be employed for the exploration of planets and their moons. Mars is the
next target for employment of rovers equipped with diverse sensors, which must necessarily use
wireless communications. The development of such vehicles is challenging since their operating
environment is not well known compared to terrestrial applications. Also, the long ranges needed
for information transmission stress the energy supplies available to such vehicles. Short-range
relay of information to a nearby station on a planet or moon, which has more stored energy, saves
the battery on a rover.
Some of the applications already discussed, especially those in buildings and transportation
systems, deal with what is called infrastructure. Systems within structures for the provision of air
and power, the handling of water and the movement of materials can all be monitored for their
integrity and performance. Highways, bridges, tunnels, ports, canals and airports are all critical to
the smooth function of the economies of societies. Other aspects of the infrastructure of a region
are also susceptible to monitoring with wireless sensor systems. These include wells, waterways,
reservoirs and piping systems for water, gas and oil. Even the integrity of sewage systems is
significant because of health problems that could attend interaction of waste with aquifers. The
surveillance of large systems for the production and distribution of power using the power grid is
also important. Monitoring of transportation and other infrastructures, and border surveillance, are
three foci of the current emphasis on homeland security in the U. S.
The locations and activities of people have already been considered regarding both health
monitoring and military operations. Other applications of wireless sensors involve systems to
monitor children, both in homes and elsewhere. Wireless systems are also being used to keep track
1
2
Neptune: A Fiber Optic Telescope to Inner Space, http://www.neptune.washington.edu/
Project GoMAP, http://mp-www.nrl.navy.mil/marine_physics_branch/gomap.htm
page-26
of the locations of prison parolees. In very general terms, the high value of people makes where
they are and what they are doing of interest for many reasons. Even valuable farm animals and
important wild animals are monitored using wireless sensor systems.
The number of potential applications of wireless sensor networks is large. They provide the
motivation for sensor research centers such as the National Centre for Sensor Research1 , and the
Center for Wireless Integrated Microsystems 2 . Basically, almost anywhere that there is something
of value or interest to anyone, it is possible to contrive a combination of sensors, microcomputers
and communications to provide information in near real time. Accessibility and power are the
primary limitations to such ubiquitous sensing. Costs vary widely, but the emergence of microtechnologies in recent years have made it possible to monitor many more quantities in many
locations for most times at affordable costs.
It is noted that three common types of inherently wireless sensors can also be impacted by microtechnologies, namely radars, lidars and sonars. Radars, which employ microwave radiation, and
lidars, which use light (laser) radiation, work in air to locate and determine the position of objects.
Small short-range radars are under development for automobile collision avoidance, detection of
people and even non-contact respiratory and circulatory measurements. These systems might
employ radio frequency MEMS. Lidars for measurement of the dynamics of structures, such as
ships and skyscrapers, can include solid-state laser sources and either point or array detectors.
Sonars are generally large, relatively low-frequency sound systems for the location of large objects
underwater. Small transducers carried on free-swimming unmanned underwater vehicles will
permit high-resolution imaging of objects in the sea. Clusters of co-located point micro-sensors that
form the nodes in a system or network are relatively new and offer great performance and promise.
Miniature versions of radar, lidar and sonar wireless sensor systems, which are on hand or under
development, are also important for volume sensing.
2.5 Medical Applications and Biodefense
Jim Wilson of Georgetown University spoke about remote sensing to assist in the handling of
epidemics. He is a member of the GDIN Infectious Diseases Working Group. He began his
presentation with a discussion of anecdotal reports of increased incidence of malaria due to both an
increase and decrease of precipitation in 1997-98. In addition, World Health Organization (WHO)
reported simultaneous outbreaks of cholera occurring in 11 countries in West and Central Africa
thought to have originated in the Horn of Africa. It was believed this outbreak was related to the
1997-8 El Nino. In March 2000, in Australia there was an MVE v. Kunjin viral encephalitis
outbreak 3 thought to be related to increased mosquito populations following heavy, sustained
rainfall. In the same month, there was an HPS (Hunta virus Pulmonary Syndrome) 4 outbreak in
Panama thought to be related to precipitation. All these linkages and others were purely anecdotal
and there was no hard evidence linking the climactic factors to the outbreaks.
1
http://www.ncsr.ie/Frame.htm
http://64.78.45.100/
3
http://www.arbovirus.health.nsw.gov.au/areas/arbovirus/viruses/murrayvalleyencephalitisandkunjin.htm
4
http://www.cdc.gov/ncidod/diseases/hanta/hantvrus.htm
2
page-27
In May 2000, in Orenburg Province, Russia, there was a CCHF (Crimean-Congo haemorrhagic
fever) 1 outbreak believed to be caused by an abrupt increase in local rodent populations thought to
be related to precipitation changes. In the same month in Walkerton, Ontario, Canada there was an
outbreak of EHEC (Enterohemorrhagic E. Coli) 2 , campylobacteria and salmonella involving an
estimated 2000 people thought to be due to flooding of sewage system of a poultry farm,
subsequently contaminating the water supply. In July of 2000 there was a single eastern equine
encephalitis fatality on the Outer Banks, of North Carolina in the US. This was attributed to
increased mosquito populations, but the link seems week. Finally, in September 2000, Israel
expressed fear that increased rainfall could result in a second wave of WNV (West Nile Viral
fever). There are similar reports of plant and livestock outbreaks of diseases.
There has been some research done in this area. Linthicum et al. looked at Rift Valley Fever in
Kenya. This is a mosquito borne disease. They used NDVI (Normalized Difference Vegetation
Index) 3 to forecast outbreaks of RVF. They found a very strong signal4 .
Colwell 5 et al. studied cholera outbreaks in Bangladesh using remotely sensed SSTs (Sea Surface
Temperatures). The epidemic was carried by copopods, which feed on algae. When the algae count
was high, copopods increased.
Glass6 studied Hantavirus using Landsat images to identify areas of high risk in the Southwestern
US. In this case, deer mice served as a reservoir and dried fecal matter was the immediate source of
infection.
Wilson and Tucker et al., studied Ebola fever. Previous outbreaks had been simultaneous, which
was very unusual and suggested climatic modulation. They used the AVHRR-NDVI (Advanced
Very High Resolution Radiometer) 7 , and established that the fever occurred in areas with
abnormally high NDVI 8 .
There have been explosive epidemics of (VEE) Venezuelan Equine Encephalitis. This can affect
humans, but its main effects are on livestock. Wilson and his colleagues are studying one outbreak
that occurred in 1995 in Venezuela and Colombia in which up to 100,000 people were infected and
300 died. The epidemic spread by up to 5 km per day. It was linked to abnormally high rainfall and
might have increased three possible vectors. Seasonal NDVIs lag precipitation by about one month.
They noted an NDVI abnormality where epidemic occurred that tracked the spread of the disease.
The readings for NDVI were among highest readings in a 21 year old dataset.
1
info at http://www.who.int/inf-fs/en/fact208.html
http://www.wvdhhr.org/bph/oehp/sdc/ecoli_protocol.htm
3
http://daac.gsfc.nasa.gov/CAMPAIGN_DOCS/LAND_BIO/ndvi.html
2
4 Linthicum, K. J. Anyamba, A., Tucker, C. J., Kelley, P. W., Myers, M. F. and Peters, C. J. (1999) Climate and Satellite Indicators to Forecast Rift
Valley Fever Epidemics in Kenya. Science, 185: 397-400.
5
http://www.umbi.umd.edu/~comb/faculty/colwell/colwell.html
http://earthobservatory.nasa.gov/Study/Hanta
7
http://edc.usgs.gov/glis/hyper/guide/avhrr
8
Tucker, C. J., J.M. Wilson, R.L. Mahoney, A. Anyamba, K.J. Linthicum, and M.F. Myers (2002) "Climatic and
Ecological Context of the 1994-1996 Ebola Outbreaks". Photogrammetric Engineering and Remote Sensing
(Special Issue on Remote Sensing and Human Health). 68(2): 147-152.
6
page-28
The work on VEE has application to US biodefense, since we have VEE vectors in the US. The
VEE is not a native virus to the US and appears to be an escapee from a laboratory. RSEPI
(Remotely Sensed Epidemic Intelligence) systems have great utility in helping to understand and
respond to epidemics, but it is very much a work in progress. Some highly experimental systems
are under development using multifactorial processes.
We conclude this section with more material supplied by David Nagel. The safety of people in
almost any setting can be improved by the use of wireless sensor networks. Monitoring the safety
of workers employed in harsh environments is already a significant application of wireless sensor
systems. Heat stress monitors are offered for sale by ten companies, some of which are wireless.
Figure 2.4 shows a commercial system capable of monitoring the conditions of as many as 10
people at ranges up to 1000 feet 1 .
Wireless sensor systems offer many opportunities for monitoring
people ranging from healthy persons, including those with
conditions such as diabetes that require continual attention, to
critically ill people in hospital intensive care units2 . Figure 2.4
shows a wireless system for monitoring temperature, heart rate
and activity. Other examples of wireless telemetry of sensor
information from the body include blood glucose levels and the
pressure within the skull and a "gut camera" (Figure 4.6). Once
the images or information are outside of the body, they can be
rapidly communicated around the globe.
Figure 2.4. Wireless
system for temperature,
heart rate and activity
sold by Mini-Mitter.
The use of wireless sensor systems to monitor elderly people in
their homes will be a rapidly growing industry in the U. S in the near future 3 . The growing fraction
of the population that is old, and the great psychological and economic benefits of keeping people
in their own homes, are two of the drivers. It is possible, and maybe even likely that the industry,
which currently provides sensors and monitors for home safety and security, will extend their
services to the monitoring of the occupants of homes. In a similar fashion, the surveillance of
persons in nursing homes can have a variety of benefits for both the occupants and their caregivers.
In hospitals, the use of wired monitors in regular and advanced care settings has been routine for
decades. In such cases, the employment of wireless systems should reduce costs and permit
reconfiguration of rooms to respond to changing requirements, such as would occur in a crisis 4 . It
is now possible to monitor the orientation, movements, heart rate, blood pressure and even
chemical factors with compact systems that can be worn by a person, and to forward the acquired
information by wireless means. Wireless monitoring of medical personnel and critical equipment
may also be done in critical situations.
1
VitalSense Physiological Telemetry System, http://www.minimitter.com/Products/VitalSense/
M. Hawley, "A Picture of Health", MIT Technology Review pp. 28-29, March 2001
3
D. Voss, "Smart Home Care", MIT Technology Review, p. 31, September 2001, and G. Sinha, "The Doctor is in the
House", Popular Science, pp.50-54, July 2000
4
G. Weiss, "Welcome to the (Almost) Digital Hospital", IEEE Spectrum, pp. 44-49, March 2002
2
page-29
2.6 Security, Crises and Military Applications1
The security of people and things, which range from valuable objects to major facilities, is a routine
requirement that can often be improved using wireless sensor networks. This has been noted in
some of the applications already discussed. Visible and infrared imagers and point sensors are
commonly used for constant security surveillance of areas and conditions. Intrusion detection is
usually the main goal of such monitoring, whether it be in homes, offices, factories, other buildings,
borders, ports or airports. Many of the systems are wired now, but wireless connectivity has some
of the same advantages for safety and security as it does for monitoring in hospitals. Cheaper
installation, easier replacement as technologies evolve and rapid reconfiguration are among the
attractions of wireless technology for security.
Crises and emergencies involve many of the same problems as routine safety and security, but with
much faster rates of change. They have many causes, as listed in Table 2.1.
Weather
Accidents, Crashes and Failures
Wind (Hurricanes, Tornadoes and Other
Transportation (Land, Marine and Air)
Storms)
Rain (Floods and Mud Slides)
Industrial
Snow (Blizzards and Deep Snow)
Structural (Collapses)
Ice Build Up
Releases (Chemical, Biological or
Nuclear)
Geological Events
Terrorism
Earthquakes and Tidal Waves
Threats
Volcanic Eruptions
Attacks
Gas Releases
Fires
Meteorite Impact
Nuclear Contamination
Fires (Forest, Urban and Others)
Nuclear Weapons
Lightening
Explosives
Failures and Accidents
Chemicals
Arson
Biologicals
Animals and Insects
Warfare
Evacuations
Many Types
Table 2.1. Natural and human causes of crises and emergencies.
The preparations for, responses to, management, mitigation, follow-up and recovery from crises,
which affect significant numbers of people, can be broadly impacted by the utility of wireless
sensor networks. Interested parties are (a) those with responsibilities, including local, regional,
national and international governments, and private companies that have contractual
responsibilities, or simply want to protect their own property, (b) those with capabilities to respond,
notably all governments, private organizations and individuals with certain skills or equipment and
(c) those affected, again including governments, companies, organizations and individuals. Some
of the activities that are germane to the various phases of crises are monitoring and surveillance;
locating and assessing problems, victims and responders; search and rescue; fire fighting; crowd
1
This section provided by David Nagel.
page-30
control; and removal or cleanup. The sensors that are useful span most of the kinds of available
sensor technologies. Pre-installed wireless systems are useful for crises because of problems with
accessibility to dangerous regions. Deployable sensors, notably those carried by small aerial
vehicles (including helicopters), are especially valuable because they can be used responsively.
A concept for a sensor-carrying wheeled robot for use during emergencies within apartment
buildings was developed recently 1 . The system would ordinarily reside in a compartment at floor
level, possibly under the niche for a fire extinguisher, on each of the floors. When commanded, or
when a fire alarm sounded, or when triggered by its own sensors, the system would move down the
hallway looking for signs of a fire using its visible and infrared imagers. It would be equipped to
wirelessly transmit images, or data from temperature or other sensors, either to managers in the
building or to firemen in the locale of the building. If widely adopted, tens of thousands of units of
such a hallway robot could find use.
There is also significant overlap between actions required by crises and the conduct of military
operations, especially warfare. A fast tempo and great uncertainty are characteristic of both crises
and warfare. Many other military functions overlap civilian applications already mentioned.
Weather is of great interest to the military for planning and executing operations, of course. The
monitoring of machinery and other systems is as much relevant to the military as it is to companies.
Perimeter monitoring and defense are equally applicable to ordinary security and to military
operations. Locating and determining the movements of friendly and hostile personnel has
similarities to following people fighting forest fires, and even some health and medical
applications. Monitoring the locations and movements of military vehicles and equipment has
features in common with following the actions of vehicles on highways.
Other functions are unique to the military. Target detection, classification, identification and
tracking are among the most compelling of military applications of wireless sensor networks.
Detection of the launch of missiles and other weapons by an enemy is of great interest. Battle
damage assessment, that is, determining the effectiveness of weapons employed by the military, is
also important. Imagers and other sensors carried on unmanned aerial vehicles are proving
especially useful for the military.
Unattended ground sensors have been developed for many years by the U. S. military. These are
wireless systems containing multiple sensors for the detection of the movements of personnel and
vehicles in their vicinity. One example is the Advanced Remote Ground Unattended Sensor
(ARGUS) system which uses an air dropped node to communicate sensed information via satellite2 .
Another system was developed by the Internetted unattended ground sensor program3 .
The U. S. Army is developing a system called the Remote Readiness Asset Prognostic/Diagnostic
System4 , to monitor the conditions within missile and munitions canisters. Temperature, humidity,
1
D. J. Nagel, unpublished, 2002.
SJW031: Advanced Remote Ground Unattended Sensor, http://www2.acc.af.mil/afc2tig/news/ newspdf/release000728-JWID.pdf]
3
DARPA's Internetted Unattended Ground Sensors, http://www.sentech-acoustic.com/page3.htm
4
Remote Readiness Asset Prognostic/Diagnostic System, http://www.dtic.mil/ndia/2001armaments/marotta.pdf
2
page-31
shock, vibration and, possibly, other factors will be monitored by a wireless hand-held interrogator.
The system is designed to have a ten-year, maintenance-free operating lifetime.
The Army is also developing a Warfighter Physiological Status Monitoring System that will report
on the condition of soldiers in the field automatically and wirelessly 1 . It has many features that are
similar to medical monitors, but will be concerned with performance as well as health matters. The
system goals are to maximize operational effectiveness while reducing casualties. Keeping track of
things, as well as people, is another focus of the military. Logistics, the supplying of all the needs
of a fighting force at the right place and time, is challenging and expensive. Systems for
identification and sensing, which report the locations and conditions of military goods and vehicles
automatically to a field or other headquarters command, are under development 2 .
Navies have a strong interest in underwater communications between ships and submarines, and
between submerged instrumentation and facilities on shore. The underwater phones and data links
that are in routine use can also be employed to transmit sensor information. While the bit
transmission rates are low compared to the use of RF in air, a great deal of data can be transmitted
over time. The situation is analogous to the transmission of pictures of distant planets and moons
from space probes. Despite the power-limited bit rates in those cases, remarkable pictures were
returned to earth.
The concept of "network centric" warfare has received much attention in recent years 3 . It springs
from the widespread availability of high-bandwidth wired and wireless connectivity, especially
using satellites. Given such communications, and the long ranges of modern weapons, it is not
necessary to co-locate sensors, weapons and personnel on the battlefield or elsewhere. Wireless
sensor systems and networks, especially but not solely imagers, are quite-naturally among the
enabling technologies for network centric warfare.
A short course entitled "Military Sensor Networks" has been offered by the Technology Training
Corporation4 . Walrod, the lecturer in that course, has provided a great deal of useful material on
wired and wireless sensor networks on the Internet and provided many useful references. 5
1
Warfighter Physiological Status Monitor: http://www.usariem.army.mil/wpsm/
Joint Total Asset Visibility, http://www.dla.mil/infoTechMain.asp
3
R. Suresh and W. E. Roper (Editors), "Battlespace Digitization and Network-Centric Warfare II", SPIE-Int'l. Society
for Optical Engineering, Volume 4741, 2002.
4
J. Walrod, "Military Sensor Networks", Technology Training Corporation, Washington DC December 2001.
2
5 Some of the sources cited these are E. Todd and J. Walrod, "Networks for Sensor Data Acquisition and Recording", THIC Meeting, Bethesda MD, 3 Oct 2000
http://www.thic.org/pdf/Oct00/psi.jwalrod.001003.pdf, ,J. Walrod, "Sensor Network Technology for Joint Undersea Warfare", NDIA Joint Undersea Warfare Technology Conference,
San Diego CA, 21 March 2002, http://www.ndia.org/committees/usw/ walrod_sensornets.pdf, and J. Walrod, "Sensor Networks for Network-Centric Warfare", Network centric
Warfare Conference, Falls Church VA, 30-31 October 2-000, http://www.plansys.com/ Content/NavigationMenu/Products/ Sensor_Network_and_Data_Acquisition_Products_
White_Papers/Sensor_Networks_for_Network_Centric_Warfare_NCW00.pdf
page-32
2.7 Recommendations
1. Much more research needs to be devoted to developing the types of sensors useful for protecting
infrastructure such as roads. There are some well –defined problems, such as vehicle framing, that
urgently need attention.
2. More education needs to be undertaken so that the nation better understands the true nature of the
terrorist threat that faces it.
3. More education about the capabilities of sensor systems needs to be undertaken in the field
science and homeland security arenas.
page-33
Chapter 3. Introduction To Field Science
3.1 Introduction
This chapter focuses on the various uses of wireless sensor networks in field science. Numerous
applications are of interest and will be discussed. The field science applications will be included
along with other outdoor applications since they have much in common. For this reason, the
various applications to homeland security discussed in Chapter 2 can suggest useful techniques for
the applications discussed in this chapter. There are several additional reviews of the use of sensors
available. 1
The talks by Bonito and Kratz presented many of the basic issues faced by field scientists. The talks
by Hughes, Nagel and Wilson also addressed some of the issues faced by field scientists. This
chapter summarizes the presentations and presents some recommendations. For videos of the
original talks and copies of the original slides, see http://homeland.cs.umaine.edu/anywhere.htm.
3.2 Remotely Deployed Sensors in Ecological Sciences
1
D. J. Nagel, "Microsensor Clusters", Microelectronics Journal, 33, 107-119, 2002, W. J. Kaiser, "Low Power
Wireless Integrated Sensors", 1994, http://www.janet.ucla.edu/WINS/lwim-innovative.htm, and D. J. Nagel, "Pervasive
Sensing", SPIE Vol. 4126, p. 71-82, 2000.
page-34
Figure 3.1. The Old Method for Retrieving Data from Sensors
Tim Kratz outlined how ecological science can benefit from remotely deployed sensors discussing
both what is currently done and what needs to be done. Figure 3.1, taken from his talk, illustrates
the old model of servicing field-deployed sensors. It provides a contrast with the scene shown in
Figure 3.2, and clearly illustrates the obvious benefits of remotely deployed sensors.
Figure 3.2. Current Remotely Deployed Sensor Systems
The National Research Council recently identified several "Grand Challenges" in the environmental
sciences. Among these, at least four can benefit from remotely-deployed sensor systems:
understanding biogeochemical cycles, biodiversity and ecosystem functioning, climate variability
and its consequences for ecosystems, and hydrologic forecasting. 1 Each of these areas is complex
and has multiple drivers acting at multiple scales. These areas are also characterized by thresholds
and non-linearities that can lead to surprises. Untangling cause and effect in these situations
requires that we have long-term observations at multiple locations, and can engage in large-scale
experimentation.
Three types of sensor systems are deployed: a) systems with a small numbers of expensive sensors,
b) systems having many moderately expensive sensors, c) systems having very many inexpensive
sensors. The state of the art for systems of type a) is more advanced than for systems of type c).
These comments are based on experience with the Long Term Ecological Research Network whose
1
National Research Council. 2001. Grand Challenges in the Environmental Sciences, National Academy Press,
Washington, DC.
page-35
stations are shown in Figure 3.3, and the 158 Members of the Organization of Biological Field
Stations 1 whose stations are shown in Figure 3.4.
Figure 3.3. The LTER Network
The most common current practice is to have a sensor attached to a data logger, with
communication via modem. One advantage is that this system works and has many protocols
developed for it. Some disadvantages are that it is expensive and does not scale well to large
numbers. In addition, it does not support direct communication among field sampling locations.
Direct communication between sensors would be very handy since it would extend the range of
communication with the sensor and would allow smart sensors to turn on and off depending on the
environmental context as detected by other sensors in the region. Because the individual units are
so expensive, it is impractical to deploy large numbers of them, and consequently it is hard to
capture environmental heterogeneity. An example of such a sensor is shown in Figure 3.5.
1
http://lternet.edu and http://www.obfs.org
page-36
Figure 3.4. The Member Stations of the Organization of Biological Field Stations in North
America.
Figure 3.5 A Relatively Expensive Sensor System
page-37
The automated buoy shown in Figure 3.5 permits the user to download stored data, and to view live
data. Communication can be two way via wireless Ethernet, so one can alter the sampling
protocols. It is easy to publish the data obtained to the WWW. A good example of a system of such
buoys is the Tropical Atmosphere Ocean Project. 1
There are also experiments that really need large numbers of sensors. Under current FCC
regulations, the spread spectrum radios currently used for communication have an effective range
of about three miles in forested landscapes, greatly limiting the spatial extent that can be monitored
effectively. Another very serious problem is the lack of power sources that are inexpensive,
dependable, rugged, and long lasting.
3.3. Instrumenting the Environment: Pervasive Environmental In-situ
Sensors
Gregory Bonito discussed in-situ sensors, which are sensors that are distributed throughout the
environment being studied. Such sensors often provide very high resolution and can provide ground
truth for satellites. Sensors can be classified into physical, chemical and biological.
Sensors are becoming smaller, faster, smarter and cheaper. In addition, their power requirements
are becoming smaller. At present over 100 properties can be sensed and there are thousands of
manufacturers.
Many government agencies, universities and commercial concerns are developing sensors. Some of
the developments are quite astonishing. One such example is the microsensor package that can be
attached to a bee, as shown in Figures 3.6 a) and b). Instrumented bees can help find explosives.
There are many other sensors that can be used on land, including sensors that can measure sap flow
non-intrusively, others that can measure various properties of soils and roots, and even ground
penetrating radar. Instrumented bees are being trained to find explosives.
There are many sensors designed for use in aquatic environments. These range from
multiparameter sondes, which can measure, more than 15 parameters, and have a datalogger2 built
in., through experimental environmental sampling processors that can analyze DNA. Other aquatic
sensor packages include acoustic Doppler current profilers, wireless moored profilers, autonomous
underwater vehicles, and digital whale tags. Common problems for aquatic sensors are fouling and
drift of the sensors.
A very promising field of development is that of MEMS3 (microelectromechanical systems).
MEMS can be used to construct very small weather stations, all types of field instruments including
single chip chemistry labs, electronic noses, electronic tongues, and small gas chromatographs.
1
http://www.pmel.noaa.gov/tao
Most of these are built by Campbell Scientific, http://www.campbellsci.com
3
http://mems.sandia.gov/scripts/index.asp
2
page-38
Figure 3.6 a) This illustration shows a bee carrying a sophisticated system of electronics on its
back. A scanner mounted on the hive downloads data when tagged bees enter and exit the hive.
Figure 3.6. b). A tagged bee showing a recent micro-telemetry system developed by Pacific
Northwest National Laboratories (PNNL). Figures a) & b) courtesy of PNNL.
page-39
Nanotechnology1 promises to make significant contributions to the field of sensors.
Nanotechnology sensors would be very small. We would expect that some would eventually get to
the atomic level, be self-repairing and possibly self-assembling.
At this time, there are many trends that can be observed. These include reducing power loads and
the use of wireless communications. There is still much work to be done on improving power
utilization. It is also clear that costs are being reduced sharply and that miniaturization will
continue. One major problem we will be facing in the near future is the flood of data that we will
get from these sensors. Figure 3.7 shows a hypothetical scene showing how one can achieve
pervasive sensing.
Figure 3.7 A Hypothetical Scene Showing Pervasive Sensing
1
http://www.nano.gov/
page-40
3.4 Communicating in the Field
David Hughes provided many details about providing communication in the wild. Much of his
report was based on the NSF-funded work he has been doing. 1 Dave feels strongly that the current
"sneaker net" that is used to transport data from dataloggers back to the home station is obsolete
and it is time to replace it with wireless.
Hughes began his talk with a tribute to the "co-inventor of spread spectrum", Hedy Lamar 2 . Spread
spectrum was classified for some 45 years. The FCC finally issued rules for using spread spectrum
in 1985. The FCC does not appear to understand the fact that spread spectrum radios can share the
same bandwidth without interference, and its rules, which have not changed since 1985, do not
make good use of the technology. Currently, spread spectrum radios are limited to 1 watt of power,
which gives them a very limited range, especially in forests, in part because of the frequencies they
employ. Also, spread spectrum is confined to three frequency bands: 902-928 MHz, 2.4-2.4835
GHz (802.11b), and 5.725-5.85 GHz. 3 Tim Shepard, in his thesis, 4 showed that it is possible for
billions of spread spectrum radios to operate in a very small area without interference. It is
important for field science and for national security that the FCC restrictions be rescinded. Such a
move is opposed by lobbyists and owners of bandwidth.
Figure 3.8 Co-Inventor of Spread Spectrum--Hedy Lamar
1
http://wireless.oldcolo.com
http://www.sss-mag.com/ss.html and http://wireless.oldcolo.com/course/hedy.htm
3
http://wireless.oldcolo.com/course/regulate.htm.
4
http://wireless.oldcolo.com/course/shepard.pdf
2
page-41
Dave Hughes has put together many sites that use wireless and sensors, and finds that the limiting
constraint is getting power to the wireless sensor units. Figure 3.9 shows his generic package.
Figure 3.9 A Generic Data Collection Package
Figure 3.10 Some Very Small Radios
page-42
Replacing the power supply in the sensor packages is now the most labor-intensive part of running
them. It is essential that some method for doing this be found. A discussion on power, and some
ideas on using lasers to supply power, are found in the next chapter. Figures 3.6 and 3.10 shows
some of the miniature radios that currently exist. To see a real-time demonstration of the sensors go
to: http://www.uaf.edu/water/projects/cpcrw/metdata/cpcrwmetsitemap.htm. For more information
see http://wireless.oldcolo.com. Figure 3.11 is a map showing one of the sites Hughes
instrumented.
Figure 3.11 The Bonanza Creek Experimental Site
3.5 Other Field Applications: Weather and Agriculture
The following material was supplied by David Nagel. There are several advantages and challenges
particular to the employment of wireless sensor networks outdoors. The major virtue is avoiding
the need for power or data cables. The main problems are provision of power to nodes and having
weatherproof systems for long-term applications. The growth of bacteria and vegetation are
problems in warmer climates. The need to operate over wide temperature ranges is another
page-43
requirement for outdoor systems. The primary outdoor applications of sensor networks are for
determination of weather, environmental parameters and factors relevant to fields of agriculture.
There are many weather parameters of interest. They include, roughly in order of decreasing
importance, temperature, pressure, wind speed and direction, humidity, dew point, rain fall amount,
air clarity, and cloud cover and height. Ordinary wireless weather systems, with individual sensors
and wireless data transmission, are a few meters in size and cost several thousand dollars.
Integrated systems, with short-range wireless links, are less than one meter in size and cost less
than one thousand dollars. 1 They provide information on most of the parameters of interest except
cloud conditions. Concepts for the development of smaller and cheaper weather stations are
available. 2
Weather measurements are needed for current conditions and for forecasting. The latter commonly
involves the use of computer models. Weather simulations are rarely done with grid points less
than one kilometer apart because of the limitations of computer performance and run times. Hence,
having meteorological measurements on grids denser than one node per square kilometer is
unnecessary now. As computers continue to increase in capability and grid spacing shrinks, it will
be desirable to have measurements on finer spatial scales. This will force the development of lowcost systems that can provide the desired measurements.
A fundamental difficulty with weather measurements and simulations is the effect of the earth's
surface. Topology and vegetation serve to strongly modify the micro-weather. Placing weather
stations on the surface, which is easy, provides only limited information on the full threedimensional weather situation. The employment of radiosondes on balloons does provide vertical
information, well away from surface effects. However, the cost of deploying balloons prohibits
their routine use. Small weather stations are likely to be mounted on existing high buildings and
towers in the future. 3
The measurement of environmental parameters goes far beyond the parameters associated with the
weather. They include the levels of water in rivers, lakes and the seas. The chemical makeup of
the atmosphere, water and soil are of interest regarding both natural and polluted conditions. The
presence and motions of pollutants in the environment, especially toxic chemicals due to spills or
willful actions, are of interest for many reasons, including legal proceedings. The number and
movements of animals are germane to biologists and wildlife managers. 4 Many different chemical
analyzers, microphones and imagers are being applied for environmental and animal
measurements. As one example, city busses in Braunschweig, Germany, are equipped with sensors
for atmosphere pollution and GPS receivers. 5 They report measured levels to a central facility via a
1
WS-2000 Wireless Weather Station, http://www.rainwise.com/ws2000/ index.html and Wireless Weather Wizard III
with Solar Power, http://www.davisnet.com/weather /products/weather_product.asp?pnum=7425WS
2
D. J. Nagel, "Microsensor Clusters", Microelectronics Journal, 33, 107-119, 2002, and M. E. Hoenk, R. Wilson and
G. Cardell, Integrated Environment Monitoring Instrument", NASA Tech Briefs, p. 34-6, November 2001.
3
Meteorological Instruments Applications Examples, http://www.youngusa.com/rmyapps1.htm.
4
A. Cerpa, D. Estrin, L. Girod, M. Hamilton and J. Zhao, "Habitat Monitoring: Application Driver for Wireless
Communications Technology", Proceedings of the First ACM SIGCOMM Workshop on Data Communications, 3-5
April 2001 and UCLA Computer Science Technical Report 200023.
5
D. Bahr, F. Schottler and C. Schlums, "Save Your Breath: GPS Drives Mobile Air Quality Monitoring", GPS World,
p.18-25, May 2002
page-44
Figure 3.12. The dots show the global distribution of
the 3000 buoys in the ARGO oceanographic sensor
system.
Figure 3.13. Photograph of a
sensor node (about three
centimeters in diameter) now
used for weather and soil
monitoring.
cellular modem. The Long Term Ecological Research stations are making increasing use of
wireless sensor systems. 1
Remote sensing of the environment is commonly done from a distance using imagers. However,
the employment of sensors in the environment that transmit their information to distant sites by
wireless means is another form of remote sensing. There are several applications of wireless
sensors in geodesy, the study of motions of the earth's surface. Oceanography, for which currents,
temperature profiles and salinities are of prime interest, is also being impacted by wireless sensor
networks. Surface studies can be done from observational satellites, and depth studies can be
accomplished by placing sensors in the seas that are read out using communication satellites. The
ARGO system is now being deployed in the earth's oceans. 2 It will consist of 3000 free-drifting
buoys that will sink to 2000 meters and then rise to transmit temperature and salinity information.
The cycle will take 10 days, with the buoys having a four to five year lifetime. The cost per buoy is
$12,000. Figure 3.12 shows the expected distribution of ARGO buoys. The collected information
will be used for climate prediction.
Weather and environmental conditions are of great interest to farmers, both for terrestrial
agriculture and for marine aquaculture. Some fields of agriculture are more sensitive to the weather
than others. A photograph of a node that measures light, temperature and humidity, and can be
equipped to do soil moisture measurements, is shown in Figure 3.13. 3 The system is being applied
to determine conditions in a botanical garden and on a farm. It was developed for planetary
monitoring by the Jet Propulsion Laboratory. The system takes measurements every second and
transmits over 40 meters. The sensor nodes are also relays, and triple hopping has been
demonstrated. Some areas of agriculture, notably forestry, have long time scales and are less
sensitive to transient conditions, but still require monitoring.
1
http://lternet.edu. A useful recent compilation of ecological sensors is available: G. Bonito, "An Ecological in situ
Sensor Resource", http://lternet.edu/technology/sensors/index.html.
2
ARGO is part of the Integrated Global Observation Strategy: http://www-argo.ucsd.edu.
3
JPL Sensor Webs Project, http://SensorWebs.jpl.nasa.gov
page-45
3.6 Recommendations
It is clear that we need cheaper, smarter sensors that can be deployed widely. It is essential that
better communication among sensors be developed. This will help distribute processing of the flood
of information that we will get. It will also extend the range at which we can collect information
wirelessly since the information can now be relayed.
It is also very important that the FCC encourage the development of spread spectrum
communications. This technology permits the sharing of spectrum and also provides more security
than conventional radio.
Finally, it is essential that reliable and cost effective methods for powering these sensor packages
be found as quickly as possible.
page-46
Chapter 4. Sensors, Wireless, and Power Sources
4.1 Introduction
This chapter provides an introduction to sensor technology. We begin the chapter with an
overview of sensors and communication technologies by David Nagel. Following that is
a section provided by Bob Nowak on power sources for sensors. We also have a section
provided by Alexander Akhmanov on using lasers to communicate with remote sensors,
and a section provided by Larry Smarr on the future of sensor systems. We conclude with
a section of recommendations.
4.2 Overview of Sensor Technology
We discussed some sensor systems possessed by humans in Chapter 1, so we will not
repeat that discussion here. The emergence of designed sensors to complement our
natural senses has given us important new tools. We can now respond to many more
quantities and to harsh conditions that would be unhealthy or lethal for humans. The
availability of wired connections, most notably the Internet, and wireless
communications, including technologies like BlueTooth, 802.11b WiFi, cellular networks
and satellite systems, give our "new" senses much greater reach and bandwidth than our
natural senses. A flood of information from sensors around the world can come to us. It
must then be coupled into our brains primarily through our vision and hearing. The
limitations of our ability to assimilate information, and the fact that we must pay attention
to our surroundings even while focused on receipt of information from sensors, have two
important implications. First, information from sensors has to be "down-sized" as much
as possible as soon as possible during its flow from sensors to humans. Consider our
vision system as a megapixel imager that receives color and brightness information from
each pixel at rates exceeding 30 Hz. It must shunt unimportant information, or else we
could not assimilate the stream of input data. So, our vision system has capabilities, such
as motion and edge detection, which permit us to focus on a small but important subset of
the inputs to our eyes. In a similar fashion, it must be possible to shed relatively useless
information in cases of high-bandwidth wired or wireless sensor networks. Second,
because information can be received from many sensors, there is the challenge of
melding that information into forms that can be consumed by humans without undo
effort. This challenge is more than "data fusion". It is really about "information fusion"
in ways that match the physiological, psychological and varied conditions and
capabilities of a wide spectrum of people. If sensor information goes to computers for
storage or use in control systems, it can be assimilated at rates exceeding what people can
handle. However, requirements for discarding useless information and for presenting it
in a proper fashion are also germane in those cases.
page-47
Since the employment of sensor systems is
about the acquisition of information, it is
PROCESSING
useful to consider the larger context
indicated in Figure 4.1. Each of the five
GENERATION
COMMUNICATION
UTILIZATION
functions of information technology shown
& Sensor Systems
& Diverse Systems
has many facets, and they evolve over time.
STORAGE
New aspects get added to each major
Figure 4.1. The major functions within
function as time passes. The employment
information technology and their
of sensors in large numbers as a means of
relationships. Sensors are rapidly growing in
generating information is a comparatively
importance as sources of information.
recent phenomenon. Ten years ago,
relatively little information came from
sensors. Ten years from now, a very significant fraction of information will originate
from sensors, especially because of the growing commercial importance of sensor
systems.
The current "Information Revolution" results from four micro-technologies: electronics,
magnetics, optics and mechanics. The connectivity between the major aspects of
information and micro-technologies is given in Table 4.1. Most of the entries are
evident. Optical computers remain under development and promise high speeds. Radiofrequency MEMS (MicroElectroMechanical Systems) are being developed for both
signal processing and communications. MEMS data storage devices, which might offer
terabit per square centimeter storage densities, are now in development. There are some
related and important aspects of information technology not shown in Figure 4.1,
including the reliability and security of computer and communications systems. However,
Figure 4.1 and Table 4.1 show the role of micro-sensors based on different technologies,
and the complex inter-relationships of information and micro-scale technologies.
µµ-Magnetics
µ-Optics
µ-Mechanics
Electronics
Generation
Computers µ-Sensors
µ-Sensors
µ-Sensors
Processing
Computers
In Development
RF MEMS
Communication Wireless
Optical Fibers & RF MEMS
Comms
MEMS
Storage
Flash
Hard Drives
Compact Discs
In
Memories
Development
Utilization
Control
Data Mining
Displays
µ-Actuators
Systems
Table 4.1. Major Processes in IT and µ-Technologies for Sensor Systems.
IT\µ-Tech:
Given the great and growing importance of the Internet, it is natural to question the
relative importance of wired and wireless sensor networks. The case for wireless sensor
networks has two primary facets. The first is the growing availability of wireless
connectivity. We are in the early stages of what has been termed the "Wireless
Revolution". BlueTooth, 802.11b WiFi, cellular networks and satellite systems have
already been noted above. The growing use of cellular systems is especially striking.
page-48
WIRELESS SUBSCRIBERS
& WIRELESS INTERNET SUBSCRIBERS
L
I
M
Figure 4.2 shows the increases in
1000
1 out of every 6 people on earth!
wireless connectivity for both voice
800
and data communications. 1 Sensors,
600
and also hand-held computers, have
400
already been integrated with cell
200
phones. It is only a matter of the
1998
1999
2000
2001
2002
2003
market before sensors, computers and
wireless communications are all Figure 4.2. The growth of wireless usage
integrated into units about the size of worldwide.
cell phones. The other driver for
wireless sensor systems and networks is the savings in costs for fixed systems, and the
need for mobility in some cases. Wires are expensive to install and protect or maintain.
Fixed wireless systems can be modified relatively easily. Issues of interference and
reliability of wireless systems, in general, and wireless sensor systems, in particular, are
germane. Usually, they depend on the specific case of interest. A recent poll by Sensors
Magazine asked at what stage respondents were in regarding implementing wireless
networking. 2 The results were: 25.6% have no plans to implement a wireless system,
17.9% were considering wireless connectivity, 20.5% had a plan for 1 to 2 years, 12.8%
had a 7-12 month plan, 7.6% would make the move in a few months and 15.3% already
have a wireless network. It is likely that the experiences of people who implement
wireless sensor networks, especially in factories and other facilities, will influence others
to do similarly, as the systems are shown to be reliable, as well as cost-effective. The
number of companies that sell and service wireless sensor systems and networks can be
expected to grow significantly in the coming years.
L
This review is motivated partly by the likelihood that experience with the use of wireless
sensor systems and networks in one industry will be useful in other industries. In
general, each major industry has its own trade shows and publications, so there is little
routine communication between large industries. Despite the great differences in the
hardware deployed and the disparate use of the gleaned information, there is a common
base of components and concerns for wireless sensor systems and networks. Such
characteristics are reviewed briefly in the next section in order to introduce the
technologies that enable and limit wireless communication of information from sensors.
I
It is necessary to make clear what is meant by a wireless sensor system and by a wireless
sensor network. Figure 4.3 illustrates the similarities and differences. Sensor systems
and networks of interest here both include sensor nodes that originate the information of
interest and users, humans or systems, which receive the information, with the wireless
transmission link in between. A system has only a single link, while a network has
multiple links, either through other sensor nodes or specific communication nodes,
depending on the architecture of the network. It is noteworthy that wireless sensor
systems will be used in conjunction with wired networks. Often the wireless system will
serve to get the sensor information to the Internet or other wired network. Sometimes,
wireless communication systems will be used to deliver the information to the users, who
1
The Industry Standard Magazine, p. 72, January 10-17 2000.
Sensors Magazine, p. 60, July 2002.
O
2
page-49
might be mobile.
In this section, we address the technologies
SENSOR
that are required for wireless sensor systems
SYSTEM
and networks. The ability to design, produce
SENSOR
and employ systems and networks depends
NETWORK
USER
on the recent and growing availability of
COMMUNICATION
several component technologies.
The
NODE
SENSOR NODE
technologies necessary for sensor nodes are
(CLUSTER)
indicated in Figure 4.4. In most cases,
Figure 4.3. Schematic of a wireless
multiple sensors will be needed to perform sensor system, and some simple
some desired function, such as the
wireless sensor networks with relayed
measurement of atmospheric conditions.
communications. In both a system and
Several sensors can beneficially share a
a network, one-way and two-way
common set of other components and power.
communications can be employed.
These include interface electronics to provide
The sensors are indicated by small
needed excitation to the sensors, and to
black rectangles.
amplify or impedance match their output
signals. The computer is commonly a micro-controller with integral analog-to-digital
converters, plus on-board program and data memories. Power can be stored in the microsensor cluster, supplied from the outside or scavenged from energy in the surrounding
region. The communication options include wired and wireless use of different
frequency carriers. The printed circuit board (PCB) offers both wiring and support for
the components of the sensor cluster. The housing provides protection, a means for
mounting and access to the ambient, when that is needed.
Microphone
POWER
SENSORS
COMMANDS
STORED
Chemical
Nuclear
INTERFACE(S)
COMPUTER
PCB & HOUSING
DATA/INFO
SUPPLIED
Electrical
RF
Acoustic
SCAVENGED
Optical
Kinetic
SEVEN
NECESSARY
COMPONENTS
CONDUITS
Electrical
Fiber Optics
Sounder
Magnetometer
1.25 in
PHYSICAL WORLD
FREE SPACE
RF
Optical
Acoustic
Figure 4.4. Schematic of a microsensor cluster showing the seven
types of components, plus the
various means of powering and
communicating.
Temperature
Sensor
Light
Sensor
Accelerometer
Figure 4.5. Photograph of the MICA microsensor cluster showing the location of the five
sensors and a sound maker (labeled sounder).
The combination of sensors, electronics, microcomputer, communications, energy source,
PCB and housing just discussed has come to be called a "micro-sensor cluster". It
frequently combines, that is, clusters, multiple sensors in order to perform functions. The
page-50
temperature, pressure, humidity and other sensors used in a weather station provide an
example. It also clusters the suite of sensors with a common base of ancillary hardware.
Mason and Wise prototyped one of the earlier clusters, which was wrist worn and
contained sensors for temperature, pressure, humidity and acceleration. 1 More recently,
Pister and his students have sought to develop micro-sensor clusters with volumes on the
order of one cubic millimeter, which they termed "Smart Dust". 2 Printed circuit boards
prototypes of such systems were developed as part of that project. 3 A new sensor cluster
from Berkeley is shown in Figure 4.5. 4 Characteristics and examples of micro-sensor
clusters have been reviewed recently. 5
A. Cluster Sensors and Electronics
Micro-sensors are attractive, both because of their performance and because of their
small size, weight and power requirements. Small sensors have been in use for decades,
with the microphone being a prime example. It was in the 1980s when micro-machined
pressure sensors appeared on the market in large numbers and in the 1990s when the
same growth happened for micro-accelerometers. The variety and number of commercial
micro-sensors that measure conditions at a location are increasing steadily now. Kaiser
recognized that the availability of micro-sensors at low cost, due to their batch
fabrication, would enable the widespread use of sensing. 6 It is true that, even if sensors
were free, the other technologies in wireless sensor networks set a floor on costs, but
commodity MEMS sensors are serving to stimulate diverse applications.
The packaging of MEMS and other micro-sensors is generally far more demanding that
the packaging of ICs. The cost of packaging commonly represents 50 to 90% of the cost
of a MEMS device. Many MEMS are in sealed packages. Indeed, there are some cases
in which ordinary microelectronic packages can be employed for MEMS. The primary
instance is the packaging of inertial sensors, including both micro-accelerometers and
angular rate sensors. However, even MEMS in sealed packages often cannot use what is
available from the IC industry. In fact, many MEMS have sealed packages that are
significantly more complicated than standard electronic packages. Optical MEMS, which
must have windows, are the primary example. Uncooled infrared sensor arrays,
consisting of pixels thermally isolated from the substrate, require vacuum packaging that
is effective for approximately a decade. Vacuum packaging is also needed for RF
MEMS that contain micro- or nano-resonators.
1
A. Mason et al, "A Low-Power Wireless Microinstrumentation System for Environmental Monitoring",
Digest Int. Conf. on Solid-State Sensors and Actuators, Stockholm, pp. 107-110, June 1995
2
B. Warneke, M. Last, B. Liebowitz and K. S. J. Pister, "Smart Dust: Communicating with a Cubic
Millimeter Computer", Computer, 44-51, January 2001.
3
S. Hollar, "COTS Dust", M. S. Thesis, Department of Mechanical Engineering, University of California at
Berkeley, 2000 (available at http://www-bsac.EECS.Berkeley.EDU/~shollar/).
4
A. Woo, "The MICA Sensing Platform", 2002, http://www.cs.berkeley.edu/~awoo/micasb.
5
D. J. Nagel, "Microsensor Clusters", Microelectronics Journal, 33, 107-119, 2002
6
W. J. Kaiser, "Low Power Wireless Integrated Sensors", 1994, http://www.janet.ucla.edu/WINS/
lwim-innovative.htm.
page-51
There are many MEMS that cannot be used in sealed packages. Pressure sensors are a
commercially important example. They are usually sealed from the ambient atmosphere
by an elastomer, which transmits pressure variations but excludes humidity and dirt.
Microphones must also be able to receive unimpeded acoustic pressure waves. MEMS
strain sensors sometimes have packages, which are penetrated by an element that is
attached to the piece being measured. Some MEMS must actually be in contact with the
atmosphere in order to function. Chemical vapor sensors are the best example. Sensors
for the analysis of water, blood and other liquids must also permit access of the exterior
environment to MEMS devices inside of packages with openings.
Maintenance of the sensitivity and calibration of gas and liquid analyzers over long
periods is challenging. This is particularly true for outdoor applications where the
temperature and humidity vary widely. Filters can be used to protect the sensitive
surfaces in analyzers that give them their functionality, but doing so requires employment
of an air mover. Even simple fans consume significant energy and draw down a battery
more rapidly. One alternative to trying to protect a functional surface, so it can be used
for extended periods, is to initially install several sensors, all but one of which are sealed.
Then, fresh sensors can be employed at preset or other times during the lifetime of the
system in the field. This tactic requires conduits, valves and a control system. Another
approach is to replace the sensor surface by moving into place a new substrate that was
stored in the overall system. This demands a mechanism that can grip and move the
substrate. Another alternative is to reconstitute the sensitive surface in place. This
requires removal of the initial thin sensor film and its reconstruction. Accomplishing
such relatively complex and demanding processes requires fairly sophisticated fluidics,
with associated reservoirs, tubing, valves, pumps and power.
Imagers are particularly useful sensors because of the dominant importance of human
vision. They can be thought of as "volume" sensors in contrast to most sensors that
provide information from a "point", that is, a specific location. In recent decades,
electronic imaging systems have improved significantly. They have also increased
dramatically both in the variety of applications and the numbers of devices in use. Their
diverse employment in security and machine vision applications are good examples. It is
noted that much of the police surveillance on the streets in Monaco is done with video
cameras, so that one person can watch many locations. There are over 1.5 million closed
circuit television systems in England, with 1400 cameras in the London Subway system. 1
Charge-coupled devices (CCDs) have diversified in opposite ways in the past decade.
Mega-pixel CCD chips have been produced in large numbers for instrumentation and,
especially, digital photography. Meanwhile, low-end CCDs have proliferated in cameras
for computers and other systems. It is now possible to buy a color and sound video
system with a CCD imager and related electronics, a transmitter, two antennas and a
1
D. Strieff, "In Britain, Smile for the Camera", MSNBC, 4 Dec 2001,
http://www.msnbc.com/news/620569.asp.
page-52
receiver for the 2.4 GHz wireless link (good to 30 meters), plus an interface box to a TV
or VCR, all for $80. 1 The imager and its electronics can be bought for half that price.
In the past few years, visible imagers based on CMOS technology have burst onto the
market. 2 They offer performance comparable to CCDs, but require significantly lower
power. One application of CMOS imagers has gotten much attention, namely their use in
a wireless imaging system for examination of the gastrointestinal tract in humans. 3
Figure 4.6 shows the pill in relation to a U.S. quarter dollar, and a cross sectional drawing
of the pill. It is noteworthy that the batteries occupy about half of the volume of the
system.
Figure 4.6. Photograph and schematic
of the endoscopic pill from Given
Imaging in Israel. Left insert: CMOS
imager is shown near a U. S. dime.
Figure 4.7. Photograph of an uncooled
infrared camera from Raytheon with a
micro-machined array of 120 by 160
pixels, and a picture taken with the
system.
Infrared imagers that detect radiation from warm objects also have changed significantly
in the recent past. Cryogenic imager arrays based on narrow bandgap semiconductors,
notably mercury cadmium telluride (HCT), have been developed for decades, primarily
for military uses. In the past decade, micromachining technology has yielded uncooled IR
imagers. While they do not have the sensitivity of HCT systems, they are substantially
simpler and about one-tenth the cost of cooled systems. Such imagers have come on the
market in the past year, with an example shown in Figure 4.7. 4 One possible application
of uncooled infrared imagers is in luxury cars for the detection of the body heat of
humans and animals beyond the range of the headlights. 5 Such cameras currently cost
about $5K, but should decline in price greatly as they come into widespread use. They
will find applications in home security, and for the monitoring of machinery and other
systems to detect hot spots that are indications of current or imminent problems.
The increasing performance and availability of microelectronics during the 1980s and
1990s was also fundamental to the development of wireless sensor networks.
Components ranging from operational amplifiers to interface chips are widely used with
sensors. Microprocessors (µP) and microcontrollers (µC) are the most important among
1
http://www.X10.com
http://www.photobit.com
3
http://www.givenimaging.com
4
Courtesy of T. Schimert, Raytheon Corporation, Dallas TX
5
http://www.cadillac.com/tech/index.htm
2
page-53
the integrated circuits that enable sensor networks. Embedded computers are used so
widely in diverse products that they are now made by 44 companies. 1 Many of those
companies offer families of computers with a broad range of capabilities, sizes and costs.
The small computers often have analog-to-digital converters (ADC) and on-chip
memory. The availability of such single-chip computers enables their co-location with
sensors in what are called "smart sensors". It is now possible to turn data, such as
voltages, into information on the parameters of interest, such as pressure or acceleration,
at the sensor by the use of a calibration curve stored in the microcomputer memory.
Compensation for extraneous factors (often temperature variations) and signal
manipulation (including averaging, integrating, differentiating and many other
mathematical functions) can be performed by the digital computer located with the
sensor. In the past, some of these functions were done electronically, essentially by the
use of analog computation. The microcomputer can also serve as the controller for
communication of the information to relay points or users. The micro-controller
functions just described are sometimes accomplished using digital signal processors or
field programmable gate arrays.
Function/Hardware Sensors Electronics
ADC
Memory
µP/µC
Transduction
Sensor Excitation
Impedance Matching
Amplification
Filtering
Signal Manipulation
Digitizing
Data/Info Storage
Apply Calibration
Apply Compensation
Control Comms
Table 4.2. Hardware in Smart Sensors that can Perform Certain Functions
The sensors and various electronics hardware just noted perform a variety of functions, as
indicated in Table 4.2. Several functions can be accomplished by some of the classes of
hardware. Microelectronics, including various integrated circuits, the ADC, memory and
the microcomputer, are involved in almost all of the functions. Some of the functions can
either be performed with analog electronics or by the digital microcomputer and
associated memory. This diversity of options offers both design flexibility and
engineering challenges.
B. Cluster Communications and Power
1
R. Cravotta, "EDN's 28th Annual Microprocessor/Microcontroller Directory", EDN, p. 33-53, 27
September 2001.
page-54
Additional functions and hardware, beyond those in Table 4.2, are needed for wireless
sensor systems and networks. They include the communication of information within the
network and, eventually, to the user(s). Radio frequency (RF), optical and acoustic links
in conduits and free space are possible, as indicated in Figure 4.4. Each has advantages
and limitations. RF links are convenient and have an immense technology base. They
can be used at rates exceeding 10 MHz. However, they generally broadcast energy into a
large solid angle. This avoids the need to aim antennas, but it uses the often-limited
energy in a sensor node at a relatively high rate. RF systems are proving to be the most
viable means for communicating information through the air from sensor nodes within a
wireless network. They include at least an oscillator to generate the carrier frequency and
means to modulate that carrier with the information to be transmitted. The latter is
generally a bit stream from the microcomputer. The use of spread spectrum techniques is
proving to be both straightforward and reliable in a wide variety of settings.
Optical systems can use substantially less energy and can transmit at rates exceeding 1
GHz. They require aiming and have weather limitations for outdoor applications. Still,
there is growing use of outdoor free-space optical links for communications of
information of any type between large buildings in cities and for connection to individual
homes. The use of modulated micro-machined corner cubes at the sensor location to
transmit information optically relieves the aiming requirement and puts the burden of
supplying energy on the interrogator. 1 Optical communication links are discussed in
Section 4.4.
Acoustic links in air, using ultrasonic carriers, can handle data rates approaching 100
kHz. However, their ranges are so short that they are generally not applicable to wireless
sensor networks. Underwater acoustic communications are heavily employed for voice
and data transmission. They are used increasingly for communication of information
from sensors. There are research, engineering and military motivations for acoustic
communication of sensor data underwater. Aside from the need for watertight and
pressure-resistant containers for both sensors and communication gear, there are
limitations and challenges associated with underwater acoustic communications. One of
them is the fact that acoustic frequencies limit the bit rates that can be transmitted.
Carrier frequencies from a few kilohertz to a few tens of kilohertz are commonly
employed underwater. Transmission rates are less than 10 Kbits per second. 2 Another
problem is the possibility of multi-path propagation of signals, which means that the
same pulse can arrive at the receiver at different times. This complicates underwater
modems. Because of scattering and background noises, the ranges of underwater
communication systems are generally limited to about 10 kilometers. Several companies
sell underwater communication links. Most of them are designed for voice
communications, and can potentially be adapted for transmission of data from sensors.
The link, whatever the carrier, between a sensor node and a user of the information from
the node is a fundamental unit of a sensor system or network. The link must have certain
1
L. Zhou, " MEMS Corner Cube Retroreflector for Smart Dust", 2000,
http://buffy.eecs.berkeley.edu/IRO/Summary/00abstracts/ lzhou.1.html
2
For examples see http://www.oceantechnologysystems.com/military.html
page-55
properties and can have others. Employment of a communication protocol is
fundamentally necessary to both successful formation of links and effective routing.
Many protocols for transmission on wired and wireless systems have been developed and
commercialized. 1 Comparisons of the features of various protocols for wireless
transmission of sensor information have been published recently. 2 Some of the protocols
available for wireless transmission of voice or data are shown in Figure 4.8, along with
their ranges and data rates. 3 Data compression methods can be used to reduce bandwidth
requirements and power consumption. In general, links that transmit images require the
greatest bandwidth, even with compression schemes such as JPEG. Frequent
measurements by point sensors, such as microphones used for acoustic detection, can also
require significant bandwidth, if there is not significant data reduction at the sensor node.
Some clusters of point sensors require relatively low bandwidth, with weather
measurements being an example.
Other aspects of the links for sensor systems and networks are important. The links in a
sensor network can have security features, such as encryption to avoid unauthorized use
of the transmitted information and authentication of a potential receiver of sensed
information. The latter requires a two-way link, with both a transmitter and receiver at
the node and at the user's location. The ability to send commands to nodes to change
their behavior, and relaying signals from one node to another, also require two-way
communication capabilities. Sensors and other nodes, which receive communications,
sometimes must have the ability to determine received signal strength in order to gauge
inter-node distances, change gain settings or trigger reconfiguration of the network. If
sensor information is delivered to mobile users, even though it might have traveled
entirely over wires until it nears the user, we strictly have a wireless sensor network.
This situation is essentially indistinguishable from the delivery of any data, such as
information from some server on the Internet, to a mobile user. It is of lesser interest
here. The general point is that wireless connectivity solves the problem of getting the
information either to the first wired node, or over the entire route, or from the last fixed
node to a mobile user of sensor information.
The provision of energy to operate the sensors, all electronics and the communications is
essentially a limiting, as well as an enabling part of wireless sensor networks. There are
many choices to be made, as indicated in Figure 4.4. The first is whether all the available
energy will be stored at the sensor locations, or whether some of it can be "harvested"
from the neighborhood of the sensors. 4 The use of solar cells for outdoor systems is the
most common example of scavenging ambient energy. Usually, it is necessary to provide
energy. This can be accomplished using nuclear sources, but doing so is rare because of
1
W. Feibel, "Encyclopedia of Networking" (Third Edition), Sybex, Alameda CA, 2000 and E. Todd and J.
Walrod, "Networks for Sensor Data Acquisition and Recording", THIC Meeting, Bethesda MD, 3 Oct 2000
http://www.thic.org/pdf/ Oct00/psi.jwalrod.001003.pdf
2
M. Wojcik, 'Wireless DAQ: Choosing the System That's Right for You", Control Solutions, p32-26,
February 2001, B. Nickerson, "Wireless LAN Technology for Engine Control and PHM", Proceedings of
the IEEE Aerospace Conference, Big Sky MT, 2002, and P. Sink, "A Comprehensive Guide to Industrial
Networks", Proceedings of Sensors EXPO, Philadelphia PA, p. 673-708, October 2001.
3
D. J. Nagel, "Microsensor Clusters", Microelectronics Journal, 33, 107-119, 2002
4
R. L. Nowak, "Energy Harvesting", http://www.darpa.mil/dso/thrust/md/energy/index.html
page-56
CDPD
CDPD
GPRS
GPRS
19.2 kbps
128 kbps
kbps
(8/900
(8/900 MHz)
MHz) (8/9
(8/9 MHz)
MHz)
1km
Range (m)
Ricochet
PHS
PHS
128 kbps
kbps
128 kbps
kbps
(900
(900 MHz)
MHz) (1.8/1.9 GHz
GHz))
802.11
802.11
&
& HRF
HRF
22 Mbps
(2.4
(2.4 GHz)
100
RFID
RFID
(13.56MHz)
10
DECT
DECT
6 4 kbps
(1.8/1.9 GHz)
GHz)
BLUETOOTH
1 Mbps
Mbps
(2.4 GHz)
GHz)
IrDA
110 kbps
kbps
IrDA
4 Mbps
Mbps
802.11b
802.11b
1 1 Mbps
Mbps
(2.4 GHz
GHz))
HiperLAN (802.11a)
5544 Mbps
(5
(5 GHz
GHz))
RadioLAN
RadioLAN
1100 Mbps
(5.8 GHz
Narrowband
Narrowband))
Cellular Digital Packet Data (CDPD)
Digital Enhanced Cordless
Telecommunications (DECT)
Global Packet Radio System (GPRS)
HomeRF (HRF)
Personal Handyphone System (PHS, Japan)
1
1K
10K
100K
1M
10M
100M
1G
Data Transfer Rate (Max Burst) bps
Figure 4.8. The physical layer data rates and ranges of wireless
networking standards. The BlueTooth and 802.11b technologies are
highlighted with boxes.
the problems attending the use of radioactive materials. Chemical sources are, and will
continue to be the primary means of powering wireless sensor networks outdoors and in
many indoor applications. In the past, batteries have been the primary choice, but small
fuel cells with about an order of magnitude more energy on volume or weight bases
compared to batteries are in prospect. Energy sources for wireless sensor networks are in
an awkward stage now. The needs for higher performance and longer lifetimes are
outstripping the slow growth in battery performance, and fuel cells are not yet available
in a variety of configurations at low costs. Commercial fuel cells for cell phones
(offering few-watt power levels for about half a day) and for notebook computers (with
about 20 watts for several hours) are in prospect by the end of this decade. The
monitoring of available energy in sensor nodes is necessary in order to control the node
and the network. Hardware (microelectronic) monitors can be used, or the on-board
computer can be programmed to estimate the energy still available. Low-power wireless
sensor systems are under development. 1
Function/Hardware
RF
System
Antenna
Energy
Source
PCB
Housing
Communications
Energy Provision
Energy Management
Electrical
Connections
Protection
1
C. G. Sondini, "The Ultra Low Power Wireless Sensor Project", http://www-mtl.mit.edu/
~jimg/project_top.html, and J. M. Rabaey, "Ultra-Low Energy Wireless Sensor and Monitor Networks",
CBS-ETAPS Workshop, Berlin, April 2000, http://bwrc.eecs.berkeley.edu/People/Faculty/jan.
page-57
Table 4.3. Additional Functions in Small Sensors
Other hardware requirements for sensor nodes in wireless networks include a printed
circuit board (PCB) to provide the electrical connectivity and a housing to protect the
node. It is noted that the antenna for the node can be printed on the PCB, and the PCB
can serve as part of the housing in some cases. Table 4.3 relates the remainder of the
hardware and functions for a sensor node. These are more uniquely correlated than is the
case for Table 4.2.
C. Network Architectures
From a network viewpoint, a micro-sensor cluster is a node in the network. Recent
advances in the technologies associated with networks have also been fundamental to
enabling wireless sensor networks. There are many choices in both the character of the
individual wireless links and the overall architecture of the network. Much work on the
design of networks falls under the aegis of computer science and technology because
computers of many sizes are commonly found in networks, and because the sophisticated
programming that is necessary for effective protocols and network functions.
The architecture of a sensor network dictates the type(s) of links that are necessary.
Designing the architecture of a network involves choices of the types and mix of nodes
for both sensing and communications, as well as their placement and possible mobility.
As indicated in Figure 4.3, there can be only sensor nodes or else a combination of sensor
and communication nodes.
Of course, other more complex and hierarchical
arrangements are possible. The sensor nodes can only transmit their information, or they
can include receivers for relaying of information or for receipt of commands. The
communication nodes may have the ability to cache and manipulate information received
from sensor nodes in their locale in order to alleviate redundant or unnecessary
communications to the user. That is, they can essentially perform as local, intermediate
dynamic servers in the network. The synchronization of clocks within wireless networks
is one of several significant requirements. The scalability of the network architecture is
often an issue. The trade-off between the overall performance, especially latency and
delay times, and the size of the network is one of the related issues. Another is the use of
reserve nodes in order to extend the overall lifetime of the network. However, that tactic
generally increases the length of communication paths, which requires more power for
communications. In general, the balance between power usage and the areal density of
nodes is a fundamental network performance and cost issue, and one of the critical design
issues.
Two of the more complex questions about the architecture of a wireless sensor network
involve its initial and subsequent organizations. In some circumstances, it is possible to
place the sensor nodes in known locations and to pre-program the organization of the
page-58
network. However, in many other applications, the sensor nodes are deployed in a
pattern that is not deterministic, and it is necessary for the network to form (self organize)
in a so-called ad hoc fashion. The nodes can be designed to determine their absolute, and
hence relative locations by incorporating Global Positioning System technology within
them. However, this increases complexity and costs of the nodes, and requires power for
the position determination. Sometimes, signal strengths can be used to infer inter-node
distances, and arrival angles can give an indication on the angles of nodes relative to each
other. In the case of fixed nodes, the position has only to be determined at the outset of
network usage. Changes in the topology of a deployed network during its lifetime can
occur due to node failures, especially due to loss of power late in the system lifetime, or
due to motion of the sensor nodes. The targets being sensed can also be mobile, so there
are four possibilities, as indicated in Table 4.4.
Fixed Sensor
Nodes
Mobile Sensor
Nodes
Fixed Targets
Time
Dependence
of
Conditions, such at the
Weather, at a Location
Finding Dangerous Objects,
such as Land Mines
Mobile Targets
People or Animals, and Objects or
Materials that are Wind or Water
Borne
People or Animals, and Objects or
Materials that are Wind or Water
Borne
Table 4.4. Examples of Sensor Use
Conservation of the energy available at the nodes is critical to the performance and
lifetime of a wireless sensor network. There are four major approaches to minimizing
energy utilization and extending the network lifetime. The first is to design nodes by
optimum choice of low-power components and circuits. 1 The second is to have the nodes
either off or in standby modes as much of the time as possible. 2 The use of network
architectures that are relatively energy efficient is the third strategy. 3 Finally, minimizing
the duration and, if possible, the range of communications consistent with the network
parameters and required performance has a large impact on saving energy. 4 Algorithms
for energy utilization by the overall network can be used to optimize various
characteristics ranging from delaying the death of the first node to maximizing the length
of time for some level of performance by the total network. There are many papers on
sensor network energy conversation because it is a central and challenging problem.
1
C. G. Sondini, "The Ultra Low Power Wireless Sensor Project", http://wwwmtl.mit.edu/~jimg/project_top.html and J. M. Rabaey, "Ultra-Low Energy Wireless Sensor and Monitor
Networks", CBS-ETAPS Workshop,
Berlin, April 2000,
http://bwrc.eecs.berkeley.edu/People/Faculty/jan.
2
L. Feeney and M. Nilsson, "Investigating the Energy Consumption of a Wireless Network Interface in an
ad hoc Networking Environment", Proceedings of InfoCom, p. 1584-57, 2001.
3
Y. Xu, J. Heidemann and D. Estrin, "Geography-Informed Energy Conservation in ad hoc Routing",
Proceedings of the Int'l. Conference on Mobile Computing and Networking, p. 70-84, 2001.
4
R. Wattenhofer, L. Li, P. Bahl and Y.-M. Yang, "Distributed Topology Control for Power-Efficient
Operation in Multihop Wireless and ad hoc Networks", Proceedings of InfoCom, p. 1388097, 2001.
page-59
D. Design and Simulation of Wireless Sensor Networks
This brief summary of the technologies that contribute to wireless sensor networks should
have made clear that the design of such networks is a complex challenge. At the node
level, the choice of components and the circuit layout is a mixed analog and digital
design problem. Designing a cluster or node in a wireless sensor network has most of the
same problems that are encountered in the design of a cell phone. Because of the fierce
competition in the cell phone industry, there are capable and complex software codes for
design of the required hardware. This is not yet the case for the design of nodes in
wireless sensor networks. Further, some wireless sensor nodes have to function in
extremes of temperature that are not encountered by cell phones in ordinary usage.
The performance of sensor nodes cannot be usefully simulated, again due to lack of
software. There are several optimization problems that could be addressed if simulation
software were available. The important trade-off between computations at a sensor node
and communications from it provides one example. Local caching of information, with
communications only at preset intervals or when some set threshold is exceeded,
increases computational energy consumption, but saves on power-hungry
communications. However, less information is available to the users of the sensor
network. There has to be an optimum mix of computation and communication that
balances energy usage against overall network utility. Facile simulation software, even at
the node level only, would contribute to determination of such optima, which are
different for different networks and even for the same network under varied conditions.
Software for the design, simulation, visualization and traffic analysis networks, given the
characteristics of the sensor nodes and networks, is available. Network simulation tools
include NS2, 1 its SenosorSim extension, 2 GloMo Sim, 3 JavaSim, 4 and OpNet 5 , among
others. Some of these are exclusively for wired networks. Networks with 100,000 nodes
have been simulated using the Scalable Simulation Framework, 6 and the
Parallel/Distributed Network Simulator7 was shown to scale to 250,000 elements.
1
S. McCanne and S. Floyd, "The LBNL Network Simulator", 1997, at http://www.isi.edu/nsnam.
A. S. S. Park and M. B. Srivastava, "Simulating Networks of Wireless Sensors", Proceedings of the 2001
Winter Simulation Conference, 2001.
3
X. Zeng, R. Bagrodia and M. Gerla, "GloMoSim: A Library for Parallel Simulation of Large-Scale
Wireless Networks", Proceedings of the 12th Workshop on Parallel and Distributed Simulations, 1998.
4
J.-Y. Tyan and C.-J. Hou, "JavaSim: A Component Based Compositional Network Simulation
Environment", Proceedings of the Western Simulation MultiConference, 2001.
5
S. Bertolotti and L. Dunand, "OpNet 2.4: An Environment for Communication Network Modeling and
Simulation", Proceedings of the European Simulation Symposium, 1993.
6
J. Cowie, H. Liu, J. Liu, D. Nicol and A. Ogielski, "Towards realistic Million Node Internet Simulations",
Proceedings of the International Conference on Parallel and Distributed Processing Techniques and
Applications, 1999.
7
G. F. Riley, M. H. Ammar and R. M. Fujimoto, "Stateless Routing in Network Simulations", Proceedings
of the Eighth International Symposium on Modeling, Analysis and Simulation of Computer and
Telecommunications Systems, 2000.
2
page-60
Simulations with millions of nodes are in prospect. 1 A taxonomy of wireless microsensor network models has been provided. 2
Since adequate software is not available for node design and simulation, it is also not
possible to rigorously co-design and co-simulate the entire node-network complex.
Neither is it possible to co-organize human and mobile robotic sensor systems, such as
might be found in crisis response teams or on the battlefield in the future. Computing
the effects of weather on wireless systems, both transmission losses and sferics due to
lightening, is challenging. Other interferences, both inadvertent and willful (jamming),
are also difficult to simulate.
E. Emplacement and Deployment of Wireless Sensor Networks
There are several ways to get micro-sensor
clusters in place for use in wireless systems
and networks. Having people install sensor
nodes has the advantage of high assurrance
of proper placement, but can be relatively
costly. Some situations will not allow
emplacement and require deployment by
some means. These include environmental,
emergency and military situations. There
are several means for such deployments.
The installation or integration of sensor
clusters with terrestrial robots offers one
deployment option. Another possibility is
Figure 4.9. Schematic of a system for
to drop sensor clusters from manned or
balloon delivery of micro-sensor
unmanned aircraft, or from missiles. Over
clusters for the detection of missile
300 unmanned aerial vehicles have been
launches.
developed for different purposes, 3 so there
are many options for micro-sensor cluster
deployment by such vehicles. Very small clusters can be dispersed by the wind using
maple seed or other bio-mimetic means for aerodynamic support. A concept for balloon
delivery was developed about a decade ago, as shown in Figure 4.9. 4 It could include
means to control the altitude of the balloon as it drifted on prevailing or other winds. The
ruggedness or micro-sensors and other components in clusters is key to their delivery by
airborne means or even artillery. 5
1
G. F. Riley and M. H. Ammar, "Simulating Large Networks: How Bib is Big Enough?", Proceedings of
the First International Conference on Grand Challenges foe Modeling and Simulations, 2002.
2
S. Tilak, N. B. Abu-Ghazaleh and W. Heinzelman, "A Taxonomy of Wireless Micro-Sensor Network
Models", to be published in ACM Mobile Computing and Communications Review (MC2R) and available
at http://opal.cs.binghamton.edu/~nael/research/papers/wcnc02-draft.pdf.
3
Jane's Unmanned Aerial Vehicles and Targets, http://www.janes.com/company/catalog/unmanned.shtml.
4
D. J. Nagel and T. J. Wieting, Naval Research Laboratory Proposal, unpublished, 1992.
5
More details are available in D. J. Nagel, "Pervasive Sensing", SPIE Vol. 4126, p. 71-82, 2000.
page-61
F. Wireless Sensor Systems and Networks
This section summarizes the goals and challenges associated with wireless sensor
systems and networks. It then provides additional information on commercial wireless
sensor nodes and existing wireless sensor systems.
Some factors, such as
manufacturability, testing and reliability that were not covered above, are recognized. A
few current programs for development of wireless sensor and related systems and
networks are noted. Potential societal and other human impacts are also cited.
Wireless sensor systems and networks share many characteristics with established fields,
especially wired industrial and other sensor arrangements, and data acquisition systems.
They are also closely related to real-time systems, and to embedded and other control
systems. Sensor networks are intrinsically distributed systems, the components of which
can be thought of as elements in a large virtual sensor system. The goals and challenges
of making wireless sensor systems and networks include most of the desires and
difficulties of designing, producing, testing and using these other systems.
The basic functions of wireless sensor systems and networks are to provide information
for awareness and decision making. They perform monitoring and surveillance,
detection, tracking and reporting towards those ends. There are many descriptors for the
specific characteristics that are sought in wireless sensor systems and networks. They
include: adaptive, responsive and reactive; autonomous, self-configuring and
coordinated; rapidly-configurable, synchronized, real-time, high-bandwidth and lowlatency; efficient and long-lived; fault tolerant, scalable and gracefully degradable;
reliable and robust; predictable, trustworthy and usable with high-confidence; distributed,
large, dense, mobile, quickly-deployable, and disposable or expendable; affordable and
cost-effective; and user-friendly.
With so many characteristics as desirata, it is clear that the design of wireless sensor
systems and networks is very challenging. It must involve many trade-offs to achieve
performance that includes a sufficient degree of the most important of the many
attributes. The challenges are complicated because the nodes, which are usually mixed
analog-digital systems, are themselves elements in the larger system or network. The
lack of all the needed simulation software for the behavior of nodes is a serious
limitation. Fabrication of the nodes is complex, but the use of common industrial
techniques based on printed circuit boards, both relieves assembly problems and makes
more reliable hardware. Testing at the node and higher levels is complex and timeconsuming. The inclusion of self testing and monitoring at those levels for behavioral
assurance is also challenging, because all such functions require more hardware or
software, and they consume additional energy. Consider the one problem of monitoring
the bit error rate of any of the many links in a network late in the life of the power
sources. Many of the challenges and applications of sensor systems and networks were
addressed in a recent retreat. 1 In short, wireless sensor systems and networks are complex
1
WEbS-Wireless Embedded Systems Winter 2001 Retreat, http://today.cs.berkeley.edu/retreat-1-02.
page-62
multi-scale engineering constructs, whose challenges are in a larger sense commensurate
with their payoffs. 1
Some commercial examples of wireless micro-sensor clusters have been cited already. In
general, a growing number of companies offer instruments with wireless communications
to a nearby computer, that is, with a link but not a network. These are termed wireless
systems, as we have noted. However, only a few companies offer nodes for use in
distributed networks. Sensoria 2 sells sophisticated four-sensor nodes, with engineering
support, for thousands of dollars each Kaiser has done experiments that involve as many
as 70 sensor nodes. 3 Crossbow 4 sells sensor nodes for hundreds of dollars each. Their
technology grew out of the "Smart Dust" program at the University of California at
Berkeley. 5 Researchers from Berkeley and Intel demonstrated a self-organizing wireless
sensor network that had over 800 nodes. 6 A plot of either the number of sensor nodes
sold, or the number used in single experiments, will undoubtedly show considerable
growth in the coming years. We expect very large numbers of sensor to be used in the
near future. 7
The area over which the sensors of any size are deployed and networked is a major aspect
of a wireless sensor network. Network descriptions such as WANs (wide area), LANs
(local area) and even PANs (personal area) are familiar. BlueTooth technology is
designed to form spontaneously what are called "piconets" within areas about 10 meters
in diameter. A relatively natural categorization of the spatial extent of networks is given
in Table 4.5.
Size Scale
1-10 meters
Terminology
Personal
10-1000 meters
Local
1-100 kilometers
Regional
Examples
An Individual's Electronic
Devices
Office
Buildings
&
Factories
Between
Buildings
&
Within Cities
Intra- & Inter-Country
Technologies
See Figure 4.8
See Figure 4.8
Cellular
Phones
Satellites
100-10K
Country
kilometers
> 10K kilometers
Global
Primarily Satellite Networks Satellites
Table 4.5. Networks Of Various Sizes and Communication Technologies.
1
R. M. Murray, "Complex, Multi-Scale Networks and Systems",
http://www.cds.caltech.edu/~murray/talks/2001b_mur01-citalum.html
2
http://www.sensoria.com.
3
W. J. Kaiser, Sensoria Corporation, private communication.
4
http://www.xbow.com.
5
B. Warneke, M. Last, B. Liebowitz and K. S. J. Pister, "Smart Dust: Communicating with a Cubic
Millimeter Computer", Computer, 44-51, January 2001.
6
Largest Tiny Network Yet, http://today.cs.berkeley.edu/800demo.
7
J. Walrod, "Sensor Network Technology for Joint Undersea Warfare", NDIA Joint Undersea Warfare
Technology Conference, San Diego CA, 21 March 2002,
http://www.ndia.org/committees/usw/walrod_sensornets.pdf
page-63
Many factors are relevant to wireless sensor networks. Prime among them is the number
and types of sensors at each node. Peak and average data transmission rates are critical to
performance. Such information is generally difficult to obtain. In the following summary,
we will be content with listing the number of sensor locations and their geographic
coverage for a few of the existing wireless sensor systems. Table 4.6 contains a
compilation of systems that involve partial or total wireless reporting from sensor nodes.
The International Monitoring System is a web of seismic stations for monitoring the
compliance with nuclear test ban treaties, and also records earthquakes. 1 The ARGO
system of 3000 buoys for oceanographic and weather measurements covers almost 70%
of the earth's surface. 2 The National Oceanic and Atmospheric Administration (NOAA)
maintains a network of moored weather buoys near the U. S. and in the South Pacific
Ocean. 3 The Tropical Atmosphere Ocean Project is another set of weather buoys near the
equator in the Pacific Ocean. 4 The traffic monitoring system on the autobahn system in
Germany uses cell phone (GSM) protocols to transmit information to the company that
installed and operates the system. 5 The Southern California Integrated GPS Network
(SCIGN) is a system of GPS receivers for detection of motion of the earth's plate in
southern California. 6 The Pacific Northwest Geodesic Array (PANGA) is a similar array
of GPS receivers that can monitor annual motions of the earth's surface as small as 5
millimeters. 7 The Texas Coastal Ocean Observation Network is a collection of 40
environmental and weather observation stations along the coast of Texas between
Louisiana and Mexico. 8 The stations communicate by satellite, packet radio and, in some
cases, telephone modem. The Puerto Rico Rain Forest Network is a set of three weather
stations in a remote national forest. 9 Many wireless sensor networks are being emplaced
to monitor and control operations in factories and plants. 10 Most of the industrial systems
will have between 100 and 1000 sensor locations in areas ranging up to about one square
kilometer, although facilities such as oil refineries will have many more sensors.
It is interesting to plot the data on the number of sensor locations and coverage for each
wireless sensor system. Figure 4.10 is the result. The vertical axis ranges from a single
sensor to 10,000 sensors. The last number might seems high, but tens times that number
of sensors is expected to be installed in coming generations of U. S. Navy ships. While
most of those shipboard sensors will still be wired, they will occupy an area much smaller
then a square kilometer. Terrestrial networks with tens of thousands of nodes are also
envisioned. The horizontal axis in Figure 10 varies from about the upper range of
1
http://www.cmr.gov.
ARGO is part of the Integrated Global Observation Strategy, http://www-argo.ucsd.edu.
3
http://www.ndbc.noaa.gov/ Maps/rmd.shtml.
4
http://www.pmel.noaa.gov/tao.
5
DDG Gesellschaft für Verkehrsdaten mbH, http://www.ddg.de.
6
http://www.sign.org.
7
http://www.geodesy.cwu.edu.
8
P. Michaud, G. Jeffries, R. S. Dannelly and C. Steidley, "Real-Time Data Collection and the Texas
Coastal Ocean Observation Network", Proceedings of the ISA Conference, Houston TX, 2001 and
http://dnr.cbi.tamucc.edu/wiki/TCOON/HomePage.
9
Hughes, private communication & http://wireless.oldcolo.com.
10
W. H. Nelson, " Survey of Wireless Technologies: Chipsets, Systems and Standards Suitable for Medium
Range, Low-Power Sensor Networks, http://www.frontier.net/ ~wadenelson/ successstories/graviton.html.
2
page-64
BlueTooth technology to the size of the earth. It can be seen that the growing use of
wireless sensor systems and networks in factories and plants will result in the highest
areal density of such technologies. Environmental networks now have about one location
in a square 1 to 10 km on a side. The German traffic network has a sensor about every 10
km. The two geodesy arrays have GPS receivers that are tens of kilometers distant from
each other on average. The ocean weather systems set up by the ARGO project, the
NOAA and the Tropical Atmosphere Ocean Project (TAOP) have sensor suites located
hundreds of kilometers apart. The International Monitoring System has stations covering
most of the globe with average sensor spacings on the order of 2000 km.
Name
Int'l. Monitoring System
Purpose
Nuclear
Explosions
ARGO Ocean Sensor Oceanography
System
NOAA Ocean Buoys
Weather
Nodes
Location
74
Global
Area
(km) 2
5.1 exp 8
3000
Global Oceans
3.5 exp 7
180
Around U. S. 9.2 exp 6
++
Pacific Equator 9.2 exp 6
Tropical Atmos. Ocean Weather
70
Proj.
DDG Network
Traffic
4000
Germany
3.5 exp 5
SCIGN
Geodesy
206
So. Cal.
3.4 exp 5
PANGA
Geodesy
6
Puget Sound
2 exp 4
TCOON
Environment/
40
Texas Coast
1 exp 3
Weather
Puerto Rica Rain Forest Environmental
3
Puerto Rico
6.4 exp 1
Many
Monitoring
& 100Factories
& 0.001 to
Control
1000
Plants
1
Table 4.6. Characteristics of Some Existing and Emerging Wireless Sensor Systems.
page-65
2
per
(10
0m
)
On
e lo
cat
ion
PANGA
2
2
per
(10
00
km
)
Puerto
Rico
Rain
Forest
NOAA
Buoys
IMS
TAOP
On
e lo
cat
ion
TCOON
On
e lo
cat
ion
km
)
per
(1
On
e lo
cat
ion
10
per
(10
km
)2
SCIGN
2
Factories & Plants
ARGO
per
(10
0k
m)
per
(10
m)
2
102
DDG
On
e lo
cat
ion
103
On
e lo
cat
ion
NUMBER OF SENSOR LOCATIONS
10 4
Area of
Earth
1
0.01
1
102
10 4
AREA COVERED (km 2 )
106
108
Figure 4.10. Areal density of nodes in existing wireless sensor systems.
Although this discussion has touched on many points, some factors very relevant to
wireless sensor networks have not been covered. Manufacturing, testing and reliability
considerations are as important for sensor networks as they are for any engineering
system. It is difficult to estimate the cost of making, putting in place and using specific
wireless sensor system now. Their variety, the early stage of their development and the
lack of appropriate software all contribute to the difficulty. It seems likely that the use of
the same technologies that are employed for the design, production and population of
printed circuit boards in consumer products will permit the costs of micro-sensor nodes to
decline with production volumes. It is certain that availability of wireless sensor
networks will both enable many new applications and reduce the costs of existing
functions, especially compared to the high costs when people perform those functions.
Testing both nodes and entire systems is challenging because of the complexity and
numbers of entities in some sensor systems. The reliability of such systems will only be
proven by field use. A decade ago, there were many questions about the reliability of
micro-machined sensors. Now that hundreds of millions of MEMS are in use, it is clear
that such micro-sensors are very reliable. Probably, the database on the reliability of
wireless sensor networks will be accumulated in a similar fashion in this and the next
decade
There are several programs for the development of wireless sensor systems and networks
in the US and elsewhere. The DARPA SensIT program includes many projects that
address various aspects of the challenges cited earlier. 1 Another DARPA program, NEST
(short for network embedded software technology) seeks to enable sensor and other
systems with up to 100,000 nodes. 2 The U. S. National Science Foundation is now
1
2
Sensor Information Technology Program, http://dtsn.darpa.mil/ixo/sensit6.htm.
Network Embedded Software Technology, http://dtsn.darpa.mil/ixo/nest.
page-66
putting strong emphasis on sensor systems and networks, both wired and wireless. Many
such systems are being developed under new homeland security programs in the U. S.
The applicability of wireless sensor networks is
sufficiently broad that the technology will
undoubtedly have implications that go beyond
what is easily foreseen. 1 For example, the
impact of ubiquitous sensing and imaging on
privacy raises many questions that bother some
people. The Center for Information Technology
Figure 4.11. Photograph of a
in the Interest of Society (CITRIS) at the
Manduca moth with the 0.7
University of California at Berkeley is
gram biotelemetry system that
addressing the technical and other aspects of
2
has two batteries and a 2 meter
major social and commercial concerns. Their
range, shown on the right.
interests include applications of wired and
wireless systems and networks in energy efficiency, transportation, environmental
monitoring, seismic safety, education, cultural research and health care.
Of all the impacts of wireless sensors systems, it may be that integration of such systems
with animals and people will prove to be the most historic. The recent demonstration of
sensing the electrical activity in the muscles of a moth in flight is an example of how
light and compact short range wireless sensor systems can be made. The insect and the
wireless sensor system are shown in Figure 4.11. 3 Even more dramatic was the
demonstrated ability of a human to control a computer by using the electro-chemical
signals that attend thinking. The initial experiment was wired, but the possible
employment of a wireless system is limited primarily by energy supply, which can be
recharged, even if implanted, by the use of RF energy. It is possible that eventually the
natural human metabolic system will be harnessed to power for implanted wireless sensor
systems. Wireless control of a rat using electrodes implanted in its brain does not use a
sensor system. 4 However, it does open the possibility of making a control loop including
both sensors and such actuation to influence behavior. One can expect that designed
sensors will be used to replace or augment our natural senses, and to provide us with
what amount to entirely new senses that couple information directly into our bodies and
1
A. L. Pennenberb, "The Surveillance Society", http://www.wired.com/wired/archive/9.12/
surveillance .html, P. Saffo, "Smart Sensor Focus on the Future",
http://www.cioinsight.com/print_article/0,3668,1=25588,00.asp, and D. Hunter, "Silent Technology-The
Next Wave", http://www.accenture.com/xdoc/en/industries/ government/Silent_technology_insights5.pdf
2
The Center for Information Technology in the Interest of Society (CITRIS),
http://www.citris.berkeley.edu/about_citris.
3
P. Mohseni, K. Nagarajan, B. Ziaie, K. Najafi, and S. B. Crary , "Robotics at the Interface of
Microsystems Technology and Biology: Biobotics", Proceedings of the Int'l. Advanced Robotics Program
Workshop on Micro Robotics, Micromachines and Systems, Moscow, Russia, p. 78-82, 24-25 Nov 1999.
4
S. Milius, "Rescue Rat: Could Wired Rodents Save the Day?",
http://www.sciencenews.org/20020504/fob3.asp and "Brain Implants",
http://www.geocities.com/skews_me/implants.html.
page-67
brains. As Kurzweil asserted, “There is no longer going to be a sharp distinction between
humans and machines". 1
4.3 Power
This section is based on Bob Nowak's presentation and other work at DARPA. 2
DARPA's work in power covers a wide range, with some sensors taking up to 100 watts
of power. A key concern is the amount of power generated for a given weight of the
system. Figure 4.12 shows what has happened regarding macro-power.
When discussing power, one must really discuss the relationship between power and
time. In particular, one must deal with such topics as peak power, average power, and the
period of time over which the power is supplied. Figure 4.13 shows some of the different
requirements for different applications. It is often hard to meet the requirements with just
one power source. For this reason, hybrid systems are widely used.
Figure 4.12 Power Densities Available Throughout History
1
R. Kurzweil", The Age of Spiritual Machines: When Computers Exceed Human Intelligence", Viking
Penguin, 1998.
2
http://www.cartech.doe.gov/pdfs/FC/129.pdf, http://www.netl.doe.gov/
publications/proceedings/01/seca/Nowak.pdf, http:// www.onr.navy.mil/ sci_tech/
chief/docs/elec_pwr_sum.ppt, and www.darpa.mil/DARPATech2000/Presentations/
dso_pdf/4NowakAdvancedEnergyB&W.pdf
page-68
For portable power systems, one should begin by seeing whether it is possible to meet the
power requirements with a battery. Batteries are hermetically sealed and very reliable,
and at this time tend to outperform most other options.
A variety of devices such as various kinds of watches can run off such things as hand
motion and body heat. This suggests to some people that there are many alternative
sources of power, but the watches only need a microwatt of power. Some of these
systems are not capable of delivering more than something on the order of a microwatt of
power, and are not suitable for more power hungry applications. In addition, people keep
demanding additional features, which constantly increases the power requirements. There
are some combinations of a PDA and a phone. One such device created by Samsung
seems to have a battery life of about two hours, which is insufficient for practical use.
In many cases, the limiting factor is not power but energy. Power is the amount of energy
per unit time. If energy is limited, that limits power, but with a lot of energy being
available, it is possible to use various storage schemes to high peak power from a source
of little power, but which produces a lot of energy over a long period of time. DARPA
and the military are focusing on energy conversion devices. These are small-scale
examples of the sorts of systems found in contemporary hybrid, electric cars. Of
particular interest are fuel cells, 1 which can convert high-energy content fuel into
electricity. Batteries and fuel cells have very similar chemistry. Batteries, however, are
closed systems, whereas fuel cells need to breathe air and need to receive fuel. For short
missions batteries are preferred because of their simplicity. For long missions, fuel cells
are preferred because of their lower weight due to the use of high-energy fuel.
The primary fuel for fuel cells is hydrogen, which can be delivered in a variety of ways.
Some recent fuel cells, developed over the last 10 years, use methanol and obtain a 30%
efficiency. 2 This gives them something like 2000 watt-hours per KG of fuel at optimal
operation. This compares very well with batteries based on their weights. Also, batteries
work best if not operated at their peak power capacity.
1
2
http://education.lanl.gov/resources/fuelcells
http://www.ott.doe.gov/pdfs/dmfuelcell3_23_01.pdf.
page-69
Figure 4.13. Power Requirements for Various DoD Applications
The military is using fuel cells with great success in the gap between batteries and large
generators. In some cases fuel cells are an order of magnitude cheaper for the amount of
energy delivered. One gallon of methanol can deliver as much energy as $4,000 worth of
batteries. People are even looking at putting fuel cells into cell phones, because they
anticipate greater energy requirements in the future. Methanol, however, has some
undesirable properties. It can be corrosive and burns with a colorless flame. Figure 4.14
shows fuel cells being used at a Marine Corps exercise. Figure 4.15 shows the relative
effectiveness of fuel cells and batteries.
Figure 4.14. Fuel Cells in Action
In some cases, flexible photovoltaics are also an option. They can be set up to charge
batteries, which can be used to power a system. Figure 4.16 shows an exercise in which a
combination of power technologies was used.
DARPA is looking at a variety of materials for direct energy conversion. These include
thermoelectric materials, among others. It is clear that we need better materials. At this
time, we want to get to 20% efficiency. The beauty of direct energy conversion is that
there are no moving parts. In addition, we want to be able to burn fuels in very small
amounts and in very clean ways to provide small amounts of heat.
page-70
Figure 4.15. Comparison of Fuel Cells and Lithium Batteries
DARPA is also looking at ways to extract energy from the environment. The best known
method of this type is photovoltaics. There are other possibilities including such things as
oxygen gradients caused by bacteria in water, methane hydrates in sediments, and the use
of other biological techniques such as the metabolism of glucose and other carbohydrates.
We are also looking at mechanical methods such as a device that might generate up to 1
watt of power from the energy absorbed by the heel in walking. 1 Another interesting
device is a mechanical eel that extracts energy from a slow moving current. 2
The energy efficiency of living organisms metabolizing carbohydrates has been under
intensive study for the past decade. One amazing feature of this energy extraction is that
it takes place at room temperature. The key to it is enzymes. There is interesting work
going on in combining enzymes with electrodes so that electrodes can use glucose. Such
a combination can be embedded in a human and provide a power source for a radio that
could be used to locate kidnapping victims.
Of great interest are various types of very small biomolecular motors. These can be very
small and very efficient, and surprisingly powerful for their size. They can be either
rotary or linear. Figure 4.17 shows some basic facts about biomolecular motors.
1
2
http://www.darpa.mil/ dso/thrust/md/energy/pa_sri.html.
http://www.darpa.mil/dso/thrust/md/ energy/pa_opt.html.
page-71
Figure 4.16. A Humanitarian Drill Using a Variety of Power Technologies
There are interesting ways to use radioisotopes to produce power. One technique is to
bombard a PN junction with radiation. Unfortunately, the radiation eventually destroys
ordinary PN junctions. Fortunately, it appears that there are now some materials that can
resist radiation indefinitely, so that radioisotopes might turn out to be a very promising
source of power.
To sum up, when deciding on a small power source, one should first consider batteries. If
batteries cannot solve the problem, one should consider small fuel cells. Next one should
consider small hybrid systems. Finally, one should consider biologically inspired
systems, or biologically based systems, although such systems are not immediately
available at this time.
page-72
Figure 4.17. A Biomolecular Motor
4.4 Laser Wireless Communications for Data Transfer from
Remote Sensors
This section is contributed by Alexander Akhmanov, who was one of the participants in
the workshop.
Free-space laser communications systems can provide a wireless communication channel
between remote sites. In these systems, the data is transmitted by modulated laser light.
Pulses of light are transmitted in a low divergence laser beam through the atmosphere.
For a variety of applications, free-space laser links have many advantages over other
wireless technologies such as RF and microwave. For example:
•
•
•
•
very high bandwidth (today's technology can achieve 2.5 GB/sec for one frequency
channel);
increased security, because of the narrow laser beam, which makes detection and
interception nearly impossible;
licensing is not required;
substantial miniaturization and integration is possible with low power consumption;
page-73
•
free-space laser terminals are mobile and are ideal for temporary installations.
There are also disadvantages:
•
•
laser free-space communications links require a clear line-of-sight;
the communications subject to atmospheric effects such as absorption of light,
refraction, scattering and scintillation.
Free-space laser communication is a proven technology. It is already used in
commercially installed systems. We estimate that the number of installed systems is more
than 10,000. There are more than 30 companies selling equipment in this market.
Free-space laser systems use transmitters and detectors that are similar to the ones used in
optical fiber communications systems, so they can achieve comparable data rates. There
are experimental installations, e.g., Lucent Technologies, with speeds ranging from 2.5 to
10 GB/sec with wavelength division multiplexing, working on distances up to 4 km.
Commercially produced systems run at rates in the range 10 Mbit/sec to 600 Mbit/sec,
and have a range of 100 m to 5 km. The most popular systems have a range of about 1.5
km.
The speed and distance are mainly depends on technical characteristics of laser source.
The majority of systems use semiconductor diode lasers (GaAs quantum well) with a
mean light power of 10 - 50 mWatts, and wavelengths of 800 - 900 nm. Some systems
use wavelengths of 1.3 - 1.5 microns common to fiber waveguides.
Optical communication systems can link spatially separated computer nets and support
all necessary protocols. It is possible to use laser free-space technology for
communications with remote sensors. The technology must be optimized for specific
applications.
For applications with remote sensors, laser and micro-optical elements for beam
conditioning can be integrated into MEMS ( MEOMS) circuits with sensor and data
processing. Such integration gives a significant reduction in size and required power.
It is possible to develop laser beam information gathering systems. In such systems, an
external laser beam reads information from sensors by detecting reflected radiation. In
such systems, there are no laser sources on the sensors. An advantage of such an
approach is the reduced power needed by the sensor package.
page-74
4.5 The Future -- A Mobile Internet Powered by a Planetary
Computer
This section is based on the presentation and slides presented by Larry Smarr. This
presentation was delivered remotely from California. This section presents some trends
that will change the way we work.
One trend is that there will be many sensors providing input to the Internet. Also, we
expect to be able to move seamlessly between technologies different technologies such as
802.11 and 3G. Substantial computing resources will be distributed all over the Earth.
This will also be combined with a substantial amount of data mining.
The Cal-IT2 Center 1 is designed to produce an integrated approach to the Internet, and to
try to look five years into the future. The Center combines 220 faculty from UCSD and
UC Irvine to work on all the subcomponents that are necessary to put together an
integrated picture of the future. There are several major areas of focus: environment and
civilian infrastructure, intelligent transportation, digitally enabled genomic medicine, and
new media arts. Since 9/11, the Center has increased its focus on homeland security. As
projects have developed, we have discovered that policy issues sometimes overshadow
technical issues.
The Defense Department is approaching many of the topics of interest to the Center in
ways similar to the way the Center is approaching them. Consequently, developments in
this area might very well constitute examples of dual use. Figure 4.18 illustrates the
Department of Defense layered architecture for defense and civilian applications.
The Center's focus is not just on terrorist threats, but on crises in general. California faces
a formidable array of natural disasters including fires, floods, earthquakes and heavy
traffic. For this reason, the Center wants to broaden the scope of homeland security to
include natural disasters.
The basic design for early warning and disaster response systems should be three tiered.
First, it should have wireless sensor nets that bring data to repositories. The second tier
would consist of collaborative crisis management centers. Finally, remote wireless
devices would interrogate databases. CAL-IT2 will focus on high performance grids like
the ones used by the ACCESS Center, as well as other grids that promote collaboration
and crisis management.
1
http://www.calit2.net.
page-75
Figure 4.18. DoD Software Architecture
Licensed and unlicensed RF bands will also add a substantial number of people to the
Internet. Figure 4.19 shows the projected growth of the Internet in terms of both the fixed
connections and the mobile users. It shows Ericsson's projections that by 2005 the
number of mobile connections to the Internet will exceed the number of fixed
connections to the Internet. This is especially surprising since it has taken something like
30 years from the creation of the ARPAnet to get the number of fixed connections in
place, but it will have taken only something like 5 years to get the mobile connections in
place. It is imperative that we make the transition between fixed points and mobile points
as smooth as possible.
G. Maria Feng’s work at the Center involves the installation of seismic sensors on
bridges. There is a lot of interest in such activities, including the commercial availability
of 802.11 “intranet” systems for bridges. A Cal-(IT)2 partner called Graviton has
developed an architecture that can connect any type of sensor via spread spectrum radio
into a network.
page-76
Figure 4.19. The Projected Growth of Fixed and Mobile Internet Connections
It looks likely that millions of video cameras will add image data streams to the Internet
in the very near future. The London Subway system has signed a contract to install
25,000 video cameras in the near future. 1 This is a system that can expand to 250,000
cameras. Surveillance cameras are also widely used on London buses. 2 It is estimated that
in London, the average citizen is seen by 300 cameras per day. Authorities are in the
process of adding face recognition to the camera network. Not everyone is happy about
this program of surveillance. 3 Washington DC is making extensive use of surveillance
cameras and planning to connect them to computers in squad cars. 4 It is clear that the
video data collected by all these cameras will add tremendously to the data stream.
Of special interest are flying platforms that can carry sensors. Figure 4.20 shows three
such platforms ranging in size from 300 inches to 1 inch.
1
http://www.telindus.com/downloads/pdf/telindus_case_london_under.pdf.
http://www.londontransport. co.uk/streets/bp_enforcement.shtml.
3
http:// www.msnbc.com/news/620569.asp?cp1=1#BODY.
4
http://loper.org/~george/trends/2002/Feb/68.html.
2
page-77
Figure 4.20. Some Flying Platforms for Sensors
It looks likely that the human body will become a multiple sensor stream source for the
Internet. For example, a company PhiloMetron1 has a smart “band-aid” that noninvasively transmits information collected from a person’s skin. It is also possible to use
invasive sensors. DARPA has been interested in the state of health of soldiers for a long
time, both in knowing what sorts of aid to send and to understand the extent of casualties.
Again, we are just beginning to think about developments in this area. This could lead to
much data and data mining.
We are in the process of really integrating sensors and wireless technology into systemson-chip. Such a step is necessary if we are to create the small and inexpensive sensor
packages that we can deploy everywhere. Such packages would include sensors,
processors, memory, protocol processors, software, and communication all in one small
package. As we know from Section 4.3, power concerns loom large, so software is
becoming very power aware and shuts down unneeded systems very quickly.
It is clear that a vast amount of software needs to be developed, creating essentially an
"operating system" for wireless sensor nets. Such a system has to be able to operate in
1
http://www.philometron.com
page-78
real time , which is quite different from the time that most operating systems operate in.
In addition, this “operating system” will have to be location aware and very secure. It
appears that cell phones will be required to be location aware within a few years, but it is
clear that adding location awareness could just be a small part of all the sensors that could
be placed on cell phones. These sensors can be chemical and biological sensors. Cell
phone companies are already working on how to get more energy into cell phones by
replacing batteries by fuel cells.
It is clear that in very short order we will have floods of data so that we need to integrate
this into what the military refers to as situational awareness. We need to view the
interrelations of variables to get a coherent picture of what is happening. The biggest
problem is getting enough pixels on a screen to display all the relevant data. Several
companies 1 are looking into this problem. Screens with something like 100 million pixels
will become common in the near future.
Figure 4.21 illustrates an existing system that integrates many functions. These include
modeling functions and prediction functions, in addition to the standard functions that
report on the status of various items. This system2 is an example of things to come. It is
also clear that training people to use such tools will become very important. Centers such
as the DC ACCESS center will play a key role in training people to work on such
systems.
An interesting development is the evolution of XML to allow the federation of
heterogeneous databases. While there are technical problems, they pale in comparison
with the data ownership problems, which revolve around deciding who can view which
database.
We now want to focus on making the wireless Internet available in your hand. It would
help the field scientist and the first responder to have the whole World Wide Web
available in the field. The Center is doing a lot of work with handheld computers
connected wirelessly. Such units are being distributed to students at UCSD, and we
expect a lot of interesting material to be developed.
We need to know how to program our cell phones, but this is difficult to do using the
native mode. Many cell phones, however, use the Qualcomm CDMA chipset, 3 which has
a high level programming environment called BREW available. 4
Figure 4.22 shows a temporary crisis center set up by Hans-Werner Braun as part of the
HPWREN 5 project. This project is getting this high-speed (45 MBS) connectivity out into
the real world, and showing how it can be used successfully.
1
e.g., http://www.panoramtech.com
http://www.saic.com/products/ simulation/cats/cats.html
3
http://www.cdmatech.com.
4
http://www.qualcomm.com/brew.
5
http://hpwren.ucsd.edu
2
page-79
Figure 4.21. An Example of a Comprehensive View of Lower NYC
We need to be overlaying physical space and cyberspace. Doing so is called augmented
reality. This suggests the more widespread use of head up displays by first responders
and a host of other people. The DoD has been working in this area for a decade and has
produced some very useful examples of heads-up displays. An interesting example of this
is creating a system that shows the stresses in a bridge when you look at it. The
widespread use of augmented reality is definitely in the cards.
page-80
Figure 4.22 Demonstrating the Use of Wireless for Crisis Response
4.6 Recommendations
We need to work on better power systems that can lower the costs associated with
maintaining a large sensor network. Existing and new technologies should be pursued in
support of portable and field communications.
We need to work on developing ideal sensor network systems. Such systems should be
modular and their components should be essentially invisible to the end user. They
should make wide use of generic components and standards so that in most cases only the
individual sensor element itself need be changed to fit a specific application.
page-81
Chapter 5. Security, Privacy and Policy
5.1 Introduction
It is clear the many important issues are not primarily technical in nature. There are
important social and political concerns as well. These range from getting people to take
security more seriously, to opening up the use of the airwaves, to financing development
in this area, and even to building communities. This chapter is based on the presentations
by Les Owens and Bill Lane, and also on contributions by Don Mitchell, Tom Williams,
Janet Thot-Thompson, Anna Komarova, Dick Morley, and George Markowsky.
5.2 Overview of Security Issues
This section is based on the presentation of Les Owens. It functions as a tutorial on
security, which is going to be of increasing importance as the use of wireless becomes
more common and the number of nodes on the Internet continues to mushroom. At the
start of his presentation, Les Owens recalled the famous quote of George Santayana:
"Those who cannot remember the past are condemned to repeat it."
Attacks on systems can be loosely grouped into two classes: passive attacks and active
attacks. Passive attacks can be further subdivided into eavesdropping and traffic analysis,
which is determining who is talking to whom even though it might not be possible to tell
what they are saying. Passive attacks are difficult to detect, but can be prevented.
Active attacks involve some intervention on the part of the attacker. Such attacks might
involve masquerading as someone else, replaying messages, modifying messages or
denial of service. These kinds of attacks are often easy to detect, but often hard to
prevent. For example, jamming is easy to detect, but might be difficult to prevent. In
some cases, detection can serve as a deterrent.
It is important to understand the sources of attacks. It is estimated that 80% of all attacks
are from insiders. This group includes malicious and disgruntled employees and former
employees. Attackers external to the organization include hackers, crackers, professional
thieves, competitors, foreign governments and terrorists.
The motivation for an attack can range from financial gain or military advantage, all the
way to doing it just for sport. Regardless of the motivation, the outcome is often the
same.
Some case studies are very instructive for better understanding security. First generation
cellular phones used to broadcast two numbers, ESN (electronic serial number) and MIN
(mobile identification number – the phone number), in addition to the number being
called. Thieves could program other phones to send the same two numbers, and some
early cell phone subscribers got colossal phone bills because calls made with the cloned
page-81
phone would all be charged on the same bill. The main motivation seems to have been
acquiring anonymity rather than stealing the phone services. Cloned phones were often
used in illegal activities such as drug smuggling and illegal arms sales.
The designers of first generation cell phones never anticipated that people would have the
capabilities of cloning phones and never prepared for it. The problem was compounded
by the fact that information on cloning phones was distributed widely through the
Internet and through other means.
There has been a great explosion in the use of 802.11 wireless LAN technology.
Unfortunately, the designers of this protocol were not security experts. In particular, the
designers did a poor job on the cryptography component, called WEP (wired equivalent
privacy), which can be broken with well-understood methods. 1 It turns out, moreover,
that people tend not to enable the security features, so it is often easy to connect to other
people’s networks. While the security features are weak, they are better than no security
at all.
It must be noted that by its nature, wireless communication is more insecure than wired
communication since it is easy to attack it passively. Spread spectrum techniques add
some measure of security, but for the best security you must use cryptography.
The fundamental requirements for security are: confidentiality (data is only accessible for
reading by authorized parties), integrity (data can be modified only by authorized
parties), authentication (you must be able to verify user’s identity), and availability (data
must be available to authorized parties).
There are additional requirements for wireless systems: interoperability (different
systems can intercommunicate), seamlessness (the ability to move from one network to
another without being dropped), and efficiency (no long waits and reasonable power
consumption).
Institutions should have a security toolbox that includes such things as firewalls, training,
and especially cryptography. Cryptography has been around for some 4,000 years and has
historically been used to protect military communications.
Classical cryptography is also referred to as symmetric cryptography, which means that
the same key (secret data) is used for encryption and decryption. If the key falls into the
hands of the enemy, the cryptographic system is compromised. For such systems, key
distribution is a problem. Public key cryptography, a relatively recent invention, has two
different keys. Either key can be used for encryption, as long as the other key is used for
decryption. In practice, you would retain one key, and can publish the other. Every
person, including any attackers, can encrypt messages, but if the cryptographic system is
secure, only the person with the other key can decrypt the message in a reasonable
amount of time. This avoids the key problem since only one key needs to be kept secret,
and only in one location. Public key systems are widely used, with the RSA system
1
http://www.cs.rice.edu/~astubble/ wep/wep_attack.html.
page-82
probably being the most widely used [http://www.rsasecurity.com/]. Public key
cryptography can be used to secure messages and to provide digital signatures, which
give a mechanism for people to send messages that make it extremely likely that the
messages came from the indicated senders.
Figure 5.1. Fixing Cellular Phone Security with Classical Cryptography
There are three layers of wireless technology. The first layer is referred to as WPAN
(wireless personal area network). This layer includes such technologies as the infrared
standard IrDA1 and the wireless standard Bluetooth. 2 The next layer is referred to as
WLAN (wireless local area network) and includes such wireless technologies as 802.11, 3
CT2 (cordless telephony second generation), UPCS (unlicensed personal communication
services), DECT (digital enhanced cordless telecommunications) and PHS (personal
handyphone system). 4 The final layer is referred to as WWAN (wireless wide area
network). This includes standards such as 3G, 5 CDPD, UMTS, 1xRTT, GPRS, GSM,
1
http://www.irda.com.
http://www.bluetooth.com.
3
http://grouper.ieee.org/ groups/802/11.
4
A good guide to these various acronyms can be found at http://www.idrc.ca/acacia/03866/wireless.
5
http://www.forum.nokia.com/html_reader/main/1,32611,2450,00.html.
2
page-83
CDMA, and TDMA. 1 There are other wireless technologies that should be mentioned:
Ricochet, RIM.Blackberry, mobile IP, and satellite.
Figure 5.1 shows how classical cryptography was used to make cellular phones clone
resistant. There are also many new ideas for making 802.11 more secure that have been
developed since the proprietary WEP protocol was developed. Cisco has a popular
approach. VPN (virtual private networks) can be used to add security to an 802.11
network. A system known as IPSec 2 is used to provide security for Internet technology,
but this protocol is too hefty for small sensor packages. We need to develop a lightweight
IPSec.
Anyone interested in security (and this should be everybody) should keep the following
thoughts in mind:
1.
2.
3.
4.
5.
Don’t underestimate the capabilities of adversaries and don’t rest on your laurels.
End-to-end cryptography is optimal – a chain is only as strong as its weakest link.
Build security into systems from the beginning.
Use good cryptography.
Design security to anticipate changes in threats and keep in mind that the action of
Moore’s Law is always providing adversaries with ever increasing computer
power.
6. Use expert help.
5.3 Privacy and Ownership of Data
The word “security” covers a broad scope, including data secrecy, data availability (i.e.,
prevention of jamming), and authentication of data. Additionally, “security” implies
conflicting interests, not only between adversaries (Good Guys versus Terrorists) but also
among allies (one’s desire for data security versus another’s desire of easy access to that
data; or the government’s need to monitor an area versus a citizen’s right to privacy) and
even internally (a scientist’s conflicting desires to provide data for others and to preserve
his own intellectual property). There are many security and privacy issues that need not
only to be balanced, but also to be juggled.
When sensors are installed for purposes of enhancing security in a facility, there are
frequently inherent conflicts with the privacy concerns of personnel that normally occupy
the sensored area. Installation of entry control devices, such as badge readers tied into a
central computer database, raise concerns that management will use the entry and exit
times to monitor employees and administer disciplinary actions. Addition of metal
detectors or x-ray screening of hand-carried items subjects personnel to a form of search
and decreases their accustomed amount of privacy.
1
Many of these terms are explained at
http://www.3gamericas.org/pdfs/Wireless_Migration_Strategy_AWS.pdf.
2
http://www.ietf.org/html.charters/ipsec-charter.html.
page-84
The solution to these concerns is at least two-fold. First, management must carefully
weigh the benefits of the additional security against the costs of those enhancements. The
costs go beyond the financial cost of the equipment purchase, installation, operation and
maintenance. Costs should also consider the increased stress placed upon workers and the
potential adverse impact on efficiency of normal activities. Second, management must
carefully explain their need for the security enhancements, then listen to and respond to
the employee concerns that will inevitably be raised.
Similarly, perhaps even more obviously, the public at large will tend towards increased
discomfort as sensors, particularly cameras, become more ubiquitous. The solution will
again involve both the balancing of security needs against other legitimate needs, and the
clear communication of those issues between the public servant and the public.
The current academic and scientific culture regarding environmental data is in conflict
with the needs of universal availability. Scientists are implicitly encouraged to keep their
data secret from others until they can assess it and publish their conclusions. This culture
is incompatible with the need to move data quickly from its gathering site to those
machines and people who are in a position to evaluate it for other uses, be they related to
homeland security, disease control, or domestic environmental monitoring (e.g., smog
levels). Thus we either need the environmental scientists to make a sacrifice, or we need
to provide incentives for them to share data.
5.4 Policy and the FCC
The environmental science and environmental security communities need to become
involved in the regulatory process regarding no-license radios. Currently, FCC
regulations regarding the no-license radio spectrum implement a homogenous rule set
throughout the U.S. Implicitly, the FCC assumes the needs of rural constituents to be the
same as those in urban areas. Such an assumption is incorrect.
The current regulatory approach is purely spectrum-based. The rationale behind this
approach is that the FCC's mission is to minimize interference by regulating the finite
resource of RF spectrum. Although such an approach has the appearance of consistency,
its execution of that approach results in inequities between rural and urban users.
Rural areas obviously have a different environment from urban ones. Rural populations
may be measured in acres per person instead of persons per acre. This difference in
environments suggests different requirements.
Rural areas are often regions of very rough terrain. Line-of-sight radio systems, such as
those operating in 900 MHz range and above, are unable to go through or around
Appalachian hills, for example. Indeed, low power radio in that frequency range cannot
even penetrate foliage very well.
page-85
Given fewer people in an area, there will be less spectrum saturation. For example, a rural
listener may receive an order of magnitude fewer stations on an FM radio than an urban
listener. The same proportions seem to hold for other segments of spectrum. Thus, there
is less to interfere with, but there are also longer distances to traverse. However, the rules
intended to minimize interference are the same whether you are in a big city or a low
population density area.
In an urban environment, a few nodes, relatively close together, can economically serve a
large number of users, be they business, residential or institutional. In a rural
environment, such economies of scale are quickly swallowed up by the long distances
between nodes. Thus, due to the low power limitations, rural users are forced to
implement expensive relay systems. This one-size-fits-all spectrum allocation algorithm
is thus heavily weighted in favor of those in urban environments. Consequently, this
spectrum allocation policy does not meet the needs of the research community, nor of the
general citizenry in low population density areas.
William Lane of the FCC spoke at the workshop on behalf of Dewayne Hendricks.
Hendricks was been called the “Broadband Cowboy” by Wired magazine in its
November 2001 issue. Lane noted that the views presented in his talk are his views and
not those of the FCC.
The FCC was established in 1934 as an independent agency of the U.S. Government,
which is responsible to Congress. It is charged with establishing a policy to govern
interstate and international communications by television, radio, wire, satellite and cable.
Note that the government’s use of communication media is handled by the National
Telecommunications and Information Administration (NTIA), which is part of the
Department of Commerce.
The FCC is headed by five commissioners, each appointed by the President of the US,
with one designated as the chair of the panel of commissioners. The FCC has seven
bureaus including the Wireless Telecommunications Bureau and the Mass Media Bureau,
and 10 offices including Engineering and Technology and Media Relations.
The Wireless Telecommunications Bureau, where Lane is the Chief Technologist, is
responsible for all domestic wireless communication except for broadcasting or satellite
communications. It provides information, licensing, rulemaking and data storage for cell
phones, paging services, personal communications services, public safety, commercial
wireless services, private wireless services and spectrum auctions.
The process for creating rules is the following. First, the FCC sends out a Notice of
Inquiry (NOI) to gather information and to generate ideas. Next, it sends out a Notice of
Proposed Rulemaking (NPRM) which includes proposed rule changes and seeks public
comment. Then it sends out a Further Notice of Proposed Rulemaking (FNPRM) to seek
further public comment on specific issues. This is followed by the Report and Order
(R&O), which issues new rules, amends existing rules or makes the decision to leave
them unchanged. These rules are entered into the Federal Register. Individuals and
page-86
organizations dissatisfied with the new rules may file a Petition for Reconsideration
within 30 days of the R&O. The FCC will then respond with a Memorandum Opinion
and Order (MO&O).
The chief goals of the FCC for spectrum management include:
1.
2.
3.
4.
Maximizing the efficient use of the radio spectrum to promote competition
Expanding access
To protect and promote the public interest
To ensure the ability of operators to adapt to new technologies, new services, and
new markets.
The basic principles espoused by the FCC include deregulation, technical flexibility,
transparency, fostering competition and holding auctions. Technical flexibility means
letting the technology develop without overburdening it with regulation. Transparency
means that the public at large is a player in developing rules and regulations.
Because of the FCC’s commitment to deregulation, the FCC always tries to ask whether
each regulation is really needed. In particular, the FCC always asks whether there are
alternatives, and whether voluntary standards are adequate. In the US, voluntary
standards are really voluntary which explains how Qualcomm was able to introduce
CDMA (code division multiple access) which is a form of spread spectrum
communication at a time that TDMA was standard for 2G. 1 Finally, the FCC always asks
whether having only one design choice is essential.
The FCC feels that technical flexibility is important since technology moves faster than
changes can be made in government regulations. Unnecessary regulations slow the
introduction of new technology by making it harder to raise capital for development. The
FCC wants the marketplace to drive the technology.
Competition is generally better than government regulation in getting new services and
technology to the public, and in lowering telecommunications costs. Figure 5.2 shows the
cell phone coverage of the United States in October of 2001. Notice that the entire
country was already covered by at least one provider, most of the country had at least two
providers, and many of the populated areas had a half-dozen service providers.
The FCC supports spectrum caps that limit the amount of spectrum that one carrier can
have as a way to foster competition. These caps are in the process of being phased out.
While advisory committees are sometimes used, public comment has a key role in
making regulations and allocations. Anyone can make comments. Public comments are
made public immediately and made available on the web. The FCC decisions explain
final rules and all the points raised in public comments.
1
For more information see http://www.cdg.org/tech/cdma_term.asp.
page-87
Figure 5.2. Cellular Coverage of the United States in October 2001
The promotion of technology is not an FCC function. The FCC has no R&D program and
is not interested in picking winners and losers, which is inappropriate for US agencies.
For example, the FCC in 1938 denied an experimental license to the first FM radio
station (W2XMN) as a “visionary development”, but it eventually granted a license and
allocated 41-44 MHz.
Congress decided in 1993 to start auctioning spectrum, and permit the market to decide
how to best use the spectrum that was purchased at auction. Purchasers of spectrum have
a lot of incentive to use the spectrum immediately so they can begin earning money.
One auction involving NextWave got complicated by a pending bankruptcy. The FCC
took back the licenses, but NextWave sued. The US Appeals Court upheld the company’s
position in June 2002 1 and the case is now pending at the Supreme Court. 2 The FCC has
an extensive website, which contains rules, 3 public comments 4 and the current frequency
allocation table . 5
1
http://www.thestandard.com/article/0,1902,27390,00.html.
http://www.usatoday.com/tech/news/ techpolicy/2002-10-08-nextwave-case_x.htm.
3
http://www.fcc.gov/mmb/ asd/bickel/47CFRrule.html.
4
http://gullfoss2.fcc.gov/cgi-bin/ ws.exe/prod/ecfs/comsrch_v2.hts.
5
http://www.fcc.gov/oet/spectrum/table/fcctable.pdf.
2
page-88
International telecommunications policy is handled with the State Department as the lead,
with assistance from the FCC and the NTIA. There are many issues with spectrum
allocation across government and non-government users, as well as differences among
nations. 3G spectrum allocation is very contentious. There are some bands, currently
unlicensed, which will be allocated at 60 GHz and 90 GHz. There are also issues
concerning ultrawideband, which has very short pulses and very wide bandwidth. The
FCC is primarily concerned about regulating power and frequency.
5.5 Technology Transfer
Sensors can be small enough to fit on a bee’s legs. They can count the axles and read the
“easy pass” of a car speeding through a tollbooth at 85 miles per hour. They can
withstand extreme environmental conditions and survive human digestion. Clearly,
sensor technology is forging ahead. But challenges exist, too. Sensors cannot power
themselves indefinitely. It is expensive and sometimes impossible to send someone to
change a sensor’s batteries. Additionally, strategically located data synthesizers are
needed in the field to collect, store, and filter the terabytes of data collected from sensors
into manageable packets that can be transmitted over existing low-bandwidth networks to
larger data processing facilities. Without wireless broadband available in remote areas to
transmit the data, the sensors and data synthesizers are of little use. These are tough
challenges that researchers are working to solve. But an equal and often overlooked
challenge is that of transferring sensor technology to wider use by practitioners and the
general public to ensure that sensor technology continues to receive support and funding
and that the technology’s benefits are maximized for the betterment of society.
This disconnect between technology advances and implementing the technology for
wider practical use is not uncommon. In their Summary of a Workshop on Information
Technology Research for Crisis Management, the Computer Science and
Telecommunications Board of the National Research Council notes the critical
interaction needed between the information technology research community and the crisis
practitioners. Without this exchange, crisis management technology and research are not
likely to have significant impact in the real world. The board asserts that among other
problems, resistance to change, insufficient education and training, and insufficient
awareness are barriers to practical use of new technology. Furthermore, sensors pose two
specific challenges. First, standards to support inter-operability among sensors must be
addressed to minimize the cost to support and adapt new sensors as they come to market.
Second, and perhaps most important, security and privacy issues must be addressed to
ensure that sensors do not invade personal lives and property or intrude into the life-styles
and cultures of people throughout the world.
Funding for research and development of sensors is largely dependent on citizens’ tax
dollars. The more that the public is able to recognize the value of sensors beyond rare
emergency situations and obscure scientific research, the more likely that dollars will find
page-89
their way to continued research, development, and production of sensors. Furthermore,
many sensors require access to wireless broadband, an expensive and regulated
commodity. The more that the public is aware of the benefits of sensors, the more likely
they will support legislative policy resulting in increased access to the wireless broadband
communications necessary for sensor functionality.
5.6 Communities
A first step in expanding outreach and education regarding sensors is identifying their
potential uses and users. Sensors can benefit the global society in the following areas:
The Military and National Security
The most easily recognizable use of sensors is that of detecting dangerous environments.
Sensors can detect chemical and biological toxins and explosives in the air and water. For
example, pervasive sensors could detect minute amounts of nerve gas or anthrax,
initiating a response and evacuation prior to a person showing signs of illness.
Crisis Preparation, Mitigation, Management, and Response
Pervasive sensors in places such as buildings, subways, trains, ships and airplanes can
allow authorities to monitor the environment at all times, quickly alerting them to
potential hazards. When a disaster (such as an earthquake, tornado or explosion) does
occur, sensors already in place can enable rescue workers to better avoid dangers and
locate survivors. Sensors at remote disaster sites permit experts located around the world
to help on-site rescuers respond to a crisis.
Distance Learning
Sensors can expand educational opportunities to rural areas or between geographically
distanced places. Video, audio, temperature, and chemical sensors, integrated with
wireless mobile communications and computers, can permit students in a class in Arizona
to observe and study penguins in Antarctica. Study of the penguins could range from
merely observing them to studying the effect of swimming on their digestion or body
temperature. Students will be able to engage in authentic science using the same data
collected by scientists at the same time to conduct their own classroom studies and
experiments. Students will be better prepared to use these same tools and techniques in
their professional careers.
Health and Medicine
Sensors enable long-distance diagnosis and medical procedures. Sensors also provide a
less invasive means of detecting internal body conditions. Sensors can be swallowed and
transmit video, temperatures and chemical data from inside the digestive tract.
Academic Research
Sensors can provide data for ecological, environmental, biological, geological and
climate research, among others. Sensor networks are able to collect and transmit data
page-90
continuously in extreme environmental locations and conditions that would be dangerous
and expensive for scientists to travel to regularly.
Technology Influenced and Enabled Genres of Art and Music
Sensors permit the integration of the environment and real-time data with art. For
example, one recent exhibit used sensors to transmit a kite’s movement through air.
Using climate sensors, a room housing an art exhibit on the African Elephant could
mimic the temperature, humidity, and air movement of the Serengeti Plains as it cycles
through days and seasons.
International and Cultural Exchange
Sensors permit a new level of cultural exchange. For example, sensors could detect the
Australian didgeridoo’s effects on airwaves and buildings in order to not only transmit
the sound, but also the feeling of the instrument to people in Uzbekistan. Body sensors
can recreate the movement sensation of dancers in Turkey for observers and learners in
Tanzania.
Remote Communication
One of the most important benefits of increased sensor use is the effect of increased
access to wireless communication capabilities for rural areas. Sensors in remote locations
are not very useful if the data cannot be transmitted to computers and scientists.
Therefore, sensor users are continually working to establish wireless broadband
capabilities in rural areas.
Commercial and Industrial
Aside from increased safety with constant environmental monitoring in industrial and
commercial environments, as sensor use grows, it creates new business opportunities and
competition in sensor development and production. This in turn makes sensor use
available for the general public.
5.7 Funding
There are several distinct areas relating to funding and support that need to be addressed.
A key area is to decide who will fund the expense of installing the various sensor systems
that will be developed to help homeland security.
This workshop sought to explore the commonality of interests between the field scientists
and homeland security community. By designing components that can be used by both
groups, we can use the power of mass production to reduce the cost of systems
tremendously.
It seems obvious that governments, at all levels, will play a major economic role in the
development of sensor networks. Of primary importance in the government funding of
these systems are the Federal agencies tasked with a role in crisis management as
page-91
evidenced by their responsibilities to the National Response Center. Also relevant are
agencies involved in weather, environmental and ecological monitoring. All of these
agencies already support the development work through their usual budgetary
mechanisms for grants and contracts, and by using innovative programs like the Small
Business Innovative Research program (SBIR). An important example of the role that the
Federal government can play is the NSF-supported High Performance Wireless Research
and Education Network. 1
Each state, through its National Guard, environmental, natural resources and other
agencies, must play a role in defining, designing, developing and deploying a persistent
monitoring infrastructure for homeland security. Local governments and independent
agencies, such as turnpike authorities, bridge and tunnel authorities, harbor authorities
and universities and research facilities, must keep homeland security issues in mind when
planning their systems. Also, the various first responder groups hold a large stake in the
development of the homeland security infrastructure.
Additionally, the private sector has a major role to play in the design, development and
deployment of pervasive sensor systems. It will probably be the private sector which, in
the long run, will actually produce and deploy the necessary systems. However, in the
shorter run, buildings and equipment, whether public or private, must be monitored
against physical, biological or chemical intrusion, and systems of various types against
wear and degradation in performance. Also, many major public (water, sewer, transit,
etc.) systems are built, and in some cases maintained and operated by the private sector,
which in such cases, assumes many of the concerns identified with local governments
(above).
For some time, steps have been underway to facilitate improved international
coordination in response to major disasters (floods, earthquakes, epidemics,
environmental disasters, etc.). One of the positive developments after September 11,
2001 has been the response of the international community to accelerating and improving
such efforts. Countries that not too recently were hostile toward each other have begun to
talk of joint security measures since it is clear that every nation is vulnerable to terrorist
attack as well as natural disasters. It seems evident that, in this context, the broad
deployment of networked sensor systems can do much to help anticipate, detect and
improve response to such problems.
5.8 Recommendations
A serious effort must be made to educate people about the importance of security and in
using security as a regular part of their communications.
1
http://hpwren.ucsd.edu.
page-92
At the same time, the technology of security must be made even more transparent and
embedded into systems so that users are not disadvantaged in using it. As part of this
effort, we need to develop a lightweight IPSec.
We need to provide incentives so that scientists who collect data in the field will share
this data with other scientists and homeland security researchers.
We need to vigorously pursue a spread spectrum bill of rights for software defined radios.
It seems clear that the effectiveness of homeland security is based on unlicensed radio,
especially in rural areas. An interesting article on this issue is Free the Airwaves! by
Kevin Werbach. 1
1
This is available on the Web in various places including
http://www.edventure.com/conversation/article.cfm?Counter=5774525
page-93
Chapter 6. Proposed Additional Workshops
We have placed conclusions in the chapters where they were most relevant. Here we
collected the recommendations for additional workshops. It is clear that the first
workshop was only a good start, and that much additional work needs to be done. To this
end, we propose that the following workshops be developed by in the near future. The
workshops are listed in order ranging from ones that are primarily technological in nature
to ones that are primarily societal and political in nature.
Workshop #1: Low Power Technology for Sensors
One of the chief constraints on deploying sensors is getting enough power to them. There
are two complementary approaches to this problem: developing better power sources and
reducing the power consumption at the sensor package. A workshop dedicated to these
issues seems of utmost importance.
Workshop #2: Wireless Sensor System Architecture
This workshop will focus on designing interoperable modular sensor systems. Some chief
design requirements will be that systems be survivable. Some specific topics that need to
be explored are the use of software and hardware agents, techniques of data gathering,
transmission, storage and archiving. Data processing and data mining are also of great
interest.
Workshop #3: Advanced Visualization Technologies
With dramatic increases in the amount of data being collected, it becomes imperative to
display this data as effectively as possible. There are many non-trivial questions to be
examined in connection with this task. One of these questions is melding sensor data with
the simulations in virtual reality environments.
Workshop #4: Security and Privacy
The increased needs for security seem to clash directly with people's desires for privacy.
It is essential to have a workshop that will focus on these issues and produce concrete
recommendations for how the new sensor-equipped environments of the future can be
designed to respect the privacy of individuals.
Workshop #5: Intellectual Property Workshop – Ownership of Data
Of great importance is the question of who owns the data that is collected by these new,
extensive sensor systems that will be built. These issues are currently being dealt with in
an ad hoc manner. It is essential that a systematic approach be taken.
page- 94
Workshop #6: Crisis Management
While our initial workshop touched somewhat on the issues of crisis management, its
main focus was on sensor systems. We recommend that there be a workshop that
examines the practical issues of crisis management from the view of local and state
governments, the Federal government, first responders, influential agencies, and nongovernmental organizations (NGOs).
Workshop #7: Outreach and Education
There is much to do in educating the research and crisis management communities in the
capabilities and uses of sensors and related systems. Such a workshop can survey
available technologies, describe their costs and performances, survey capital and
operational costs, and cover legal and societal factors relating to the use of such
technologies.
page- 95
Appendix A
Biographies of Speakers
Gregory Bonito is an ecosystem scientist with a focus on soil and biogeochemical
processes. Currently, he is Research Assistant with the LTER (Long Term Ecological
Research) Network Office and is involved in creating a resource/database on
environmental in situ sensors.
Capt Dennis "Mike" Egan is a graduate of the USCG Academy (1972). He completed
the Adm. Rickover program at MIT (1977) with an MS in Mechanical Engineering and
the Professional Degree of Ocean Engineer. He graduated from the Naval War College
(CNCS) and Salve Regina College in 1987 with MA degrees in Strategic Analysis of
National Security and International Relations respectively. In 1991 he graduated from the
Sloan School of Management at MIT, as a Sloan Fellow with a MS degree in
Management with emphasis on Negotiation.
He served on six different Arctic and Antarctic deployments and was the Engineer
Officer of the icebreakers Northwind and Polar Star. During a follow-on assignment as
the Naval Engineer of the Second Coast Guard District in St. Louis, MO he was
recognized as the Coast Guard's Federal Engineer of the Year and received the Federal
Energy Award for innovative design and construction of new, integrated Coast Guard
barges and new modular configuration towboats that were the first river tenders to carry a
mixed gender crew.
Following graduation from the Naval War College, he served as the Coast Guard's
Liaison to the Republic of the Philippines during the first two years of the Aquino
Administration (1987-1989) and was posted at the Joint US Military Assistance Group as
the J-2 and Support Division Officer. He served as Vessel Support Branch Chief at
MLCLANT (1989-1990) and was appointed as a Sloan Fellow (MIT). He served as a
member of the Commandant's Strategic Planning Staff and was detailed to the staff of
Vice President Al Gore as a Special Assistant to the Director of the National Performance
Review (NPR) (1992-1994), for which service he received the Legion of Merit.
He served as Commander, Group Honolulu and Base Commanding Officer of CG
facilities at Sand Island, Honolulu. During his tour (1994-1996), his units distinguished
themselves in several major AMIO cases, inter-island counter-narcotics cooperative
actions with state and local authorities, and over 500 search and rescue cases. He was
transferred to the Seventeenth Coast Guard District as the Chief of the Maritime Plans
and Policy Division where he also coordinated regional Russian and Canadian programs.
In addition, he served as the Alaska Regional Response Team Co-Chairman, Exercise
Director of Exercise Northern Edge (1997-1999), the First Spill of International
Significance Exercise in Sakhalin, Russia (1998) and Chairman of the D17 Intelligence
Coordination Council. He transferred to Commandant (G-OPF) in August 1999. Captain
Egan is a registered Professional Engineer and a Lifetime member of the US Naval
Institute.
page-96
During the periods of Y2K and the disastrous events of 9-11, he served as the Chief of he
Coast Guard's National Command Center and the National Response Center, the nation's
first alert and notification center for weapons of mass destruction and terrorist attack. He
served as the Coast Guard's senior contingency planner during the weapons of mass
destruction national exercise TOPOFF and was a co-founder of the Multi-Sector Crisis
Management Consortium, bridging the world of research and cutting edge
communications and information technology with the world of the practical emergency
responder. He recently retired from the Coast Guard after over 34 years and now is Chief
Scientist and Director of Homeland Security Intermodal Transportation at the System
Planning Corporation in Arlington, VA.
Rich Holm holds a B.S .in electrical engineering and M.S. in nuclear engineering from
the University of Illinois. He served in the United States Navy as a nuclear trained
Machinist Mate/Engineering Laboratory Technician for eight and a half years. During
that period he was an Engineering Watch Supervisor aboard a nuclear submarine and an
instructor in chemistry, radiological controls, mechanical and hydraulic systems at a
nuclear prototype unit. He has worked as a consultant to Northern States Power
Company and the International Atomic Energy Agency developing training programs in
health physics and nuclear engineering.
Mr. Holm is currently the reactor administrator for the University of Illinois Nuclear
Reactor Laboratory. He is responsible for all aspects of the operation and safety of the
University of Illinois Advanced TRIGA and LOPRA nuclear reactors including the
following: compliance with Technical Specifications and Federal regulations, license
amendments, safety evaluations, equipment/system design changes and experiment
review and authorization. He also performs special projects for the College of
Engineering including coordinating activities for the University of Illinois Program for
Security Technology.
Dave Hughes is a partner of Old Colorado City Communications, a Colorado Springs,
Colorado, USA Internet ISP, which does research and consulting in advanced forms of
wireless data communications. His company has been located since 1984 in the Old
Colorado City District of westside Colorado Springs.
Hughes is a West Point graduate, Class of 1950, who served in the Korean and Vietnam
wars, was a researcher and staff officer in the Office of the Secretary of Defense in the
1960s. He retired as Colonel, US Army, in 1973. He has been researching, developing,
and operating, online dial up, or wirelessly connected systems since 1979. He has
assembled a variety of communication systems in rural areas such as the Big Sky
Telegraph system in Montana in the late 1980's.
In 1998 Wired Magazine named Hughes one of the leading "Wired 25" pioneers in the
world. He consulted for the Congressional Office of Technology Assessment in 1990 and
more recently advised the Federal Communications Commission (FCC) use of wireless
for education under the 1996 Telecommunications Act. He is a lecturer worldwide. He
page-97
published most recently in the April 1998 issue of Scientific American and the MIT Press
'First Hundred Feet' compendium in May 1999.
In 1993 Hughes was awarded the Telecommunications Pioneer Award in by the
Electronic Frontier Foundation (EFF) of Washington, DC for his effective work in grass
roots connectivity. He informally assisted the Puerto Rican Office of Management and
Budget when it started its Wireless Project in 1997. He has been an invited speaker at the
Puerto Rican Governor's Conferences on Information Technologies in 1998 and 1999.
For the past 4 years he has been the Principal Investigator for 4 National Science
Foundation projects doing Field Tests of advanced wireless technologies for education
particularly no-license spread spectrum devices.
He designed a novel Field Science by Wireless Project for educators in Lewistown
Montana, in 1997-98, and directed the implementation of Wireless connectivity for
educational and science institutions in Ulaanbataar, Mongolia in 1995-96.
He has been sought out after as a lecturer and conference panelist worldwide for the past
20 years. He is currently Principal Investigator for a 3 year NSF project that will model
Field Science by solar powered wireless connections from remote sensors in frigid
climates and tropical rain forests to the Internet, in support of biological sciences and
environmental monitoring.
Tim Kratz is a Senior Scientist at the Center for Limnology at University of WisconsinMadison and Director of the Center’s Trout Lake Station in northern Wisconsin. He is an
ecosystem ecologist with interests in the long-term landscape ecology of lakes, landwater interactions, and lake biogeochemistry. He is a Co-Principal Investigator of the
North Temperate Lakes Long-Term Ecological Research project. Over the past several
years he has been involved with developing and deploying instrumented buoys to
measure various physical, chemical, and biological dynamics of lakes.
William Lane is currently the Chief Technologist of Wireless Telecommunications
Bureau for the Federal Communications Commission.
Since January 1999, Mr. Lane was the Chief Scientist with Femme Comp Incorporated
and served on the staff of the Director for Information Systems for Command, Control,
Communications and Computers (DISC4) at Headquarters, Department of the Army,
where he was responsible for the Joint Tactical Radio System Program.
Previously, he completed a career as a US army Signal Corps Officer culminating with
the rank of Colonel and assignment as the Deputy Head of the Department of Electrical
Engineering and Computer Science at the United States Military Academy. His
assignments and responsibilities included a broad range of tactical communications
ranging from special operations to division and corps level communications as well as
strategic level communications with the former Defense Communications Agency. In
addition, he served as an Instructor in the Electrical Engineering Department at the
United States Military Academy and as the Special Projects Officer for the Chief of the
Army Section of the Joint United States Military Mission for Aid to Turkey.
page-98
He received his Ph.D. from the Georgia Institute of Technology, his MBA degree from
Long Island University, and his BS degree from the United States Military Academy. He
is a senior member of the Institute for Electrical and Electronics Engineers (IEEE) AND
IS A registered Professional Engineer in the state of Virginia.
George Markowsky is currently Professor and Chair of the Computer Science
Department at the University of Maine. He is also the Chair of the Mathematics &
Statistics Department. George Markowsky has published 69 papers on various aspects of
Computer Science and Mathematics. He has written an additional 19 technical reports
and 6 books on various aspects of computing. He also holds a patent on mechanisms that
implement Universal Hashing. His interests range from pure mathematics to the
application of mathematics and computer science to biological problems.
He has also built voice controlled and enhanced keyboard terminals for use by paralyzed
individuals. George Markowsky received a grant from NSF to found the Agent Institute
at the University of Maine to focus research on agent-based computing. He is also a co-PI
on the University of Maine’s Internet 2 grant. George Markowsky served as the President
of the Maine Software Developers Association (MeSDA) since its inception in spring
1993 until May 1998. The Association has grown from 18 members to over 200 and now
employs a full-time Executive Director. MeSDA works closely with companies and state
agencies to promote the development of the software industry in Maine.
George Markowsky founded a software company, Trefoil Corporation in February 1994.
Trefoil Corporation developed the O*NET software for the U. S. Department of Labor
that will replace the Dictionary of Occupational Titles. The O*NET software was
released nationally in 1998. In addition, Trefoil has handled tasks ranging from software
reengineering and testing, to the marketing of a product called PC-Pedal™. Trefoil is
currently working on a Phase II SBIR for the National Institutes of Health.
Since that time he has founded several companies, including Ayers Island, LLC, which is
developing a research/commercialization complex two miles from the university of
Maine on a 63-acre island, and Maine Venture Capital. He served as Secretary of the
Multi-Sector Crisis Management Consortium from 2001 through 2003. For details about
the projects cited above and others check: http://www.umcs.maine.edu/~markov.
David J. Nagel graduated (magna cum laude) from the University of Notre Dame (B.S.
in Engineering Science 1960), and performed graduate work at the University of
Maryland (M.S. in Physics 1969, and Ph.D. in Engineering Materials 1977). During
active duty with the Navy, he was Navigator aboard the USS ARNEB on OPERATION
DEEPFREEZE (1960-2), and then served as a Technical Liaison Officer at the Naval
Research Laboratory (NRL) (1962-4). After joining the civilian staff of the NRL in 1964,
he held positions of increasing responsibility as a Research Physicist, Section Head,
Branch Head and, finally, Superintendent of the Condensed Matter and Radiation
Sciences Division. In the last position, he was a member of the Senior Executive Service,
and managed the experimental and theoretical research and development efforts of 150
page-99
government and contractor personnel. At the NRL, Dr. Nagel’s research interests
centered on radiation physics, especially x-ray spectroscopy, and on materials sciences,
with applications to materials analysis, plasma diagnostics, integrated circuit production,
environmental studies, "cold fusion", and MicroElectroMechanical Systems (MEMS).
He has written or co-authored over 150 technical articles, reports, book chapters and
encyclopedia articles. He is lead-author of a patent on x-ray lithography, which formed
the basis of a 100-person startup company in Rochester NY.
After serving as Commanding Officer of three Reserve units and the national Technology
Mobilization Program, Dr. Nagel retired as a Captain in the United States Naval Reserve
in 1990. He left Government Service, and became a Research Professor in the School of
Engineering and Applied Science of The George Washington University, in 1998. He is
now working on the development and applications of MEMS and microsystems for the
military and other sectors, with special attention to radio frequency and acoustic systems.
Robert Nowak managed advanced energy technologies projects aimed at reducing the
logistics, weight, and volume burden of military power sources while lowering acoustic
and thermal signatures for the Defense Advanced Research Project Agency (DARPA).
Advanced battery and capacitor technology, mobile electric power, portable power, and
energy harvesting projects have been and are supported by this program. Rechargeable
lithium batteries and capacitor efforts have been transitioned to the Army and Air Force
for further development for specific mission needs. The mobileMElectric Power Program
succeeded in producing fuel processing technology that can process high sulfur content
diesel and jet fuels for use in large (tens of kilowatts) fuel cell systems. This program has
been transitioned to the Navy, who, in cooperation with the Coast Guard and other
agencies, is developing fuel cell systems for shipboard use. Recent projects are portable
power and energy harvesting. The portable power effort seeks to develop highperformance, logistically fueled power sources in the range of hundreds of watts. For
lower power ranges, other fuels, such as methanol and novel hydrogen sources, are being
integrated with small fuel cells; however, logistics fueled systems are preferred by the
military and are the long-term goal of these programs. The Energy Harvesting Program
seeks new concepts to reduce or eliminate the dependence on batteries in smallunattended sensors or soldier systems by accumulating energy from environmental
sources for immediate or deferred use. Prior to his tenure with DARPA, Dr. Nowak was a
program manager at the Office of Naval Research and managed the Navy's basic research
program in electrochemistry as well as undersea propulsion programs. He was a staff
scientist and section head at the Naval Research Laboratory. He received postdoctoral
training at the University of North Carolina, Chapel Hill, a doctorate in chemistry from
the University of Cincinnati, and bachelor's and master's degrees from Oakland
University in Rochester, Michigan.
Leslie D. Owens is currently employed as the Technical Director, Wireless Security for
the Global Professional Services group of Nortel Networks. In this role, he is responsible
leading the development of global wireless security solutions sets and for security support
to ensure solution sets offer security capabilities within the Wireless Internet segment.
page-100
Prior to joining GPS, Mr. Owens was Principal Security Architect / Strategist and
Director, Network Security Services within the Enterprise Solutions organization. He was
responsible for defining the cross-product security strategy and security implementation
in the Nortel enterprise voice and data products. He worked with security architects and
product line managers in the various product groups to determine consistent, robust, and
interoperable security features to ensure the end-to-end security of Nortel products. Prior
to joining Nortel, Mr. Owens was responsible for the strategic development and
implementation of the overall enterprise-wide network security and fraud management
program for Iridium LLC, its gateways and service providers worldwide. Before Iridium,
Mr. Owens established and led AT&T Wireless Services’ nationwide, cross-divisional
network security efforts. For 5 years, at GTE Laboratories, Mr. Owens was involved in
applied cryptography and security research for GTE’s wireless and wireline business
units, directed the cellular industry’s fraud laboratory and chaired the US digital cellular
security group. Earlier experience includes quick-reaction digital logic and
microprocessor design work at the National Security Agency. Mr. Owens has been
involved in network security and fraud control for more than 15 years. He has published
and spoken both nationally and internationally on network security and fraud control and
holds four patents and has 4 patents pending on inventions for fraud control,
cryptography and network security. Mr. Owens is a writer and editor for Wireless
Security Perspectives (www.cnp-wireless.com) and is an Adjunct Professor in Computer
Science at Georgetown University. Mr. Owens holds a BS in Electrical Engineering from
Virginia Tech, an MS in EE from George Washington University, and completed all
except the dissertation for a Ph.D. in EE from Northeastern University.
Larry Smarr was the founding director of NCSA and the National Computational
Science Alliance, while a professor of astronomy and physics at the University of Illinois
at Urbana-Champaign.
In August 2000, he moved to the University of California San Diego, where he is a
professor of computer science and engineering. In December 2000, he became the
founding director of UC's California Institute for Telecommunications and Information
Technology, which has a focus on the software, wireless, and photonic technologies
necessary for extending the capabilities of the Internet into a global grid. He is a member
of both the President's Information Technology Advisory Committee and the Advisory
Committee to the Director of NIH. He is also a Member of the National Academy of
Engineering, and a Fellow of the American Physical Society and the American Academy
of Arts and Sciences.
Jim Wilson of Georgetown University is a pediatrician with extensive experience with
infectious diseases and remote sensing (e.g., correlation of climatic and vegetation
changes with outbreaks of disease such as Ebola, Rift Valley Fever and VEE).
page-101
Appendix B
Biographies of Participants
Alexander Akhmanov is Head of the department and Senior Scientist in the Institute on
Laser and Information Technologies, Russian Academy of Sciences ( ILIT RAS). (1999 to present date).
Bill Bellamy is a former member of the EPA Drinking Water Advisory Committee and is
a recognized expert in water treatment technology and policy issues. He is also an expert
in disinfection, which is a significant aspect of water system security, as well as in
chemical contaminants. Bill is very knowledgeable about the specific agents that are of
concern from a drinking water perspective, the effectiveness of current water treatment
technologies in removing such constituents, and system vulnerabilities.
Michael R. Chritton RISS ESH&Q Manager from CH2M HILL.
Kevin Fall, Research Computer Scientist at Intel Research lab, Berkley. You can see
some of the work at http://tinyos.millennium.berkeley.edu/ .
Kevin Gamble - ADEC (American Distance Education Consortium). He is a non-profit
ISP for universities – especially research scientists - connecting to the Internet via
Satellites – such as Tachyon systems. He spoke at the Organization of Biological Field
Stations in September, with lots of info on costs, technical challenges and choices.
http://www.adec.edu/satellite-resources.html.
Brad Hutchens is a Senior Project Manager/Environmental Engineer in the Directorate
of Health Risk Management of the US Army Center for Health Promotion and Preventive
Medicine (USACHPPM). He is primarily responsible for environmental surveillance
for military deployments and is the Program Manager for European Command (EUCOM)
environmental surveillance. Currently he directs environmental method development
initiatives for military deployments at USACHPPM and has developed the “Deployment
Environmental Surveillance Database” which has become the template for the
Department of Defense Occupational Health Readiness System (DOHRS) environmental
module. Mr. Hutchens’ past professional experience includes directing environmental
field surveys, and sampling and evaluating of military sites. This included environmental
sampling in support of Ballistic Missile Defense incentives. He is also experienced in
conducting environmental compliance inspections and has conducted environmental
compliance and sampling training. He has a bachelor’s degree in Civil Engineering from
Virginia Polytechnic Institute and State University and a master’s degree in
Environmental Engineering from Johns Hopkins University. He is a registered
Professional Engineer in the State of Maryland.
Anna Komarova is Professor of Linguistics, Head of the Department of Foreign
Languages for Geography (1996- to present date) - Lomonosov Moscow State
University.
page-102
Richard E. Morley is best known as the father of the programmable controller and is the
leading visionary in the field of advanced technological development. He is also an
entrepreneur whose consistent successes in the founding of high technology companies
has been demonstrated through more than three decades of revolutionary achievements.
Mr. Morley is the recipient of the Franklin Institute's prestigious Howard N. Potts Award
and is an inductee of the Automation Hall of Fame. He holds more than twenty US and
foreign patents, including those for the parallel inference machine, the hand-held
terminal, the programmable logic controller and magnetic thin film.
Using his studies in physics at the Massachusetts Institute of Technology as a
springboard, Dick Morley has become an internationally recognized pioneer in the fields
of computer design, artificial intelligence, automation and technology trend forecasting.
As an inventor, author, consultant and engineer, Dick Morley has provided the Research
and Development community with world changing innovations.
For many years, Dick Morley was a contributing columnist to Manufacturing Systems
Magazine. He has also written articles for magazines and journals worldwide including
Manufacturing Automation Magazine. In recognition for his ground breaking
contributions, Mr. Morley has received numerous awards and honors from such diverse
groups as Inc. Magazine (Entrepreneur of the Year), and the Society of Manufacturing
Engineers (Albert M. Sargent Progress Award)
Dick Morley lives on a farm in New Hampshire and works out of a renovated barn on his
property. He has raised over two dozen children and loves skiing. He drives a 1995
Chevrolet Impala SS and rides a Harley-Davidson Sturgis motorcycle.
John Porter is a Research Assistant Professor of Environmental Sciences at the
University of Virginia. He works with the Virginia Coast Reserve Long-Term Ecological
Research Project, which, with the help David Hughes and Tom Williams, has established
an Internet-connected wireless network linking research, sites on Hog Island (8 miles off
the Eastern Shore of Virginia). For additional information, see the project web site at:
http://www.VCRLTER.virginia.edu.
Janet Thot-Thompson, Director ACCESS Center. Janet is the National Center for
Supercomputing Applications Associate Director for the Alliance Center for
Collaboration, Education, Science and Software (ACCESS). She helped conceptualize,
design, deploy, and now manages this advanced technology and learning center in
Arlington, Virginia for the National Computational Science Alliance (the Alliance). The
Center opened April 1999. See: http://calder.ncsa.uiuc.edu/ACCESS/
Additionally, she is the Acting Executive Director for the Multi-Sector Crisis
Management Consortium (http:// www.mscmc.org). The MSCMC is a 501-C (3) nonprofit tax-deductible organization.
page-103
ACCESS is a high-end computing, communications and information technology center
for technology demonstration, presentation, training, digital video collaboration and
distance learning Center for the Alliance. The center helps showcase advanced and
emerging software, hardware, and communications technologies from across the
Alliance. The National Computational Science Alliance (Alliance) is a partnership among
more than 50 academic, government, and industrial organizations from across the United
States to prototype an advanced computational infrastructure for the 21st century. This
model infrastructure, called the Grid, will link together advanced supercomputers,
visualization environments, and mass storage devices into a powerful, flexible problemsolving environment.
Janet has more than 29 years of experience managing customer service organizations in
both the private and public sectors and as a director for two technology centers. She was
an active participant with the U.S. National Performance Review to reinvent government
through advanced technologies, specifically the Reinvention Laboratory, RegNet, and has
given numerous presentations before professional associations. Software Publishing
Corporation recognized her work in March 1999 in the Washington Post and in
Government Computer News in March 1991 for innovative, systematic application of
software. She was also recognized in Federal Computer Week in numerous articles as
the founding member and Chair of the World Wide Web Federal Consortium during
1995-1996. Ms. Thot-Thompson received the second highest honor granted by the
Nuclear Regulatory Commission to an individual, the Meritorious Service Award,
received for Management Excellence. She was a panel member at the National Press
Club March 23, 2000 "cyber cocktail" on the future of the Internet. She coauthored (with
Mary Bea Walker, NCSA Senior Research Scientist and Karen S. Green, NCSA
Assistant Director of Communications) a white paper on Innovative Collaborative
Learning and Research Environments in Academia and Government: Developing the
NCSA ACCESS Center. Most recently she consulted with the Asian Development Bank
(ADB) in Manila, Philippines, and presented on advanced technologies and use of
potential applications to reduce poverty. As a result, ADB is budgeting to join the MultiSector Crisis Management Consortium and also for an ACCESS-like advanced
technology center.
As Acting Executive Director for the Multi-Sector Crisis Management Consortium, she is
dedicated to harnessing the power of supercomputing, communications, and advanced
applications and visualization development to enhance international crisis management
prevention, mitigation, response and consequence management activities.
Michael Wartell is a chemistry Professor and the chairman of the Defense Intelligence
Agencies Science Board.
page-104

report - Dr. George Markowsky

Transcription

Similar documents

Wireless Sensor Troubleshooting Guide

iSCOUT PRS80

MACHINE CONDITION MONITORING FOR A CARIBBEAN RESORT

CargoLink™ Wireless Sensors

Two forces working with the weather…

get protected! - Pine Cellular

THERMOMATIC EC HOME®

- Saurashtra University

ASSURANCE - Sensaphone

The idea with modifying stock down pipes is to gut the pre