Mobile Systems: Annual Report 2006

Transcription

Annual Report 2007
Institute of Networked and Embedded Systems
Professor Christian Bettstetter
Professor Mario Huemer
Professor Bernhard Rinner
NES.uni-klu.ac.at
1
Publisher:
Alpen-Adria-Universität Klagenfurt
Lakeside Science and Technology Park B02
9020 Klagenfurt, Austria | Europe
Web: http://nes.uni-klu.ac.at
E-Mails: <firstname.lastname>@uni-klu.ac.at
Photos:
Bernhard Horst, Bernhard Lamprecht, Wolfgang Schriebl, Michaela Wassner
IT-Campus Kärnten, Ulrich Neffe, Gerhard Maurer, Infineon Technologies and private photos of staff members
Layout: Christian Bettstetter
Typesetting: Heidelies Aschbacher
Software:
Adobe® InDesign® CS3
Printed by:
Carinthian Bogendruck GmbH, Klagenfurt
Number of printed pieces: 1.000
© 2008 Alpen-Adria-Universität Klagenfurt
2
Table of Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
People . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Teaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Activities and Administration . . . . . . . . . . . . . . . . . . . 63
3
Introduction
Dear research partners, colleagues, and students, dear friends:
I invite you to celebrate with us the first birthday of
the Institute of Networked and Embedded Systems!
Its foundation in January 2007 demonstrated the
university’s commitment to a new emphasis in the area
of information and communication technologies.
Since then many things have happened. Two new
research groups have been established—the “Pervasive
Computing Group” and the “Embedded Systems and
Signal Processing Group.” My two new colleagues,
Prof. Bernhard Rinner and Prof. Mario Huemer, did
a great job in establishing their groups, acquiring first
research funding and working out several courses. The
institute grew from 6 to 23 permanent staff members
and scholars. Many new lectures and courses for the
new curriculum Information Technology have been
offered, and also the number of enrolled students grew
to almost one hundred.
This report gives an overview of our institute: the
people, their research and teaching activities, and some
social aspects illustrating our working atmosphere.
The research activities are grouped around the topics
algorithms and protocols, architectures, networking
theory, signal processing, and hardware-oriented
issues, with a focus on wireless and mobile networks of
4
pervasive devices and sensors. Our expertise includes
cooperative wireless communications, networked
cameras, synchronization and localization, selforganization in networks, and sensor fusion.
In the year 2007, two new externally-funded projects
started, namely
•
•
the “EVis project” on autonomous traffic monitoring
and
the “SPARSE project” aiming at exploiting
mobility in wireless networks.
Additional projects have been prepared and will start
soon:
•
•
•
•
the “RELAY project” on new techniques for
cooperative communication in wireless networks,
an FWF-funded project on communication
architectures for robot systems,
the “PowerDCDC project” on concepts and circuits
for power management units funded by our local
industrial partner Infineon Technologies Villach,
the “VHD project” on very high data rates for near
field communication funded by NXP Gratkorn,
and
•
a project on digital signal processing concepts for
RFICs funded by DICE Linz and LCM Linz.
A major milestone of the year 2007 was the
establishment of Lakeside Labs—a universityaffiliated research platform with a technical focus
on self-organizing networked systems. The technical
and organizational orientation of the lab has been
developed by all three professors of the institute. It
will serve as an incubator for the ICT institutes of
the university, providing about 3.5 Million EUR per
year for an initial period of five years. In particular, it
will enable larger projects with up to seven researchers
working on visionary and multidisciplinary projects.
With involvement of several institute members, we
defined a scientific and organizational framework
and are currently in the stage of starting first research
projects.
A lot of our time has been invested in teaching,
covering bachelor, master, and PhD courses for
the study programs Information Technology and
Informatics. New undergraduate lectures include the
second semester course on Electronic Circuits and
the two third semester courses on Systems Theory
and Communication Engineering and Design of
Digital Circuits. At the graduate level, several new
lectures in the areas of pervasive computing and signal
processing as well as a new lab course on embedded
microcontrollers have been established.
Furthermore, we continued our series of guest lectures.
One of the teaching highlights has been the lecture
Communication Theory in Mobile Systems and
Genetics by Prof. Joachim Hagenauer which was very
well received by our students.
Preparations for a new, very comprehensive ICT lab
course, in which students will gain valuable hands-on
experience, are underway.
Professors celebrating the first birthday of NES
Several staff members were (and still are) engaged
in committees of the university and in the research
community. As two highlights, I would like to mention
the First ACM/IEEE International Conference on
Distributed Smart Cameras organized by the Pervasive
Computing Group and the election of Prof. Rinner as
vice dean of our department.
Last but not least, the institute’s technical infrastructure
has been extended significantly, including a new server
pool for high-performance simulations, large network
storages, and lab equipment.
Enjoy this report, and thanks to all who make this
institute what it is!
Christian Bettstetter
Professor and Institute Head
5
Our Mission
6
The Embedded Systems and Signal
Processing Group focuses on system-,
algorithmic-, architectural- and hardware-oriented aspects of embedded
systems. Application areas are short
range and cellular wireless systems,
wireless positioning systems as well
as automotive applications. Our vision is to work on long-term research
projects as well as on short-term topics, the latter in close cooperation
with industry partners. In teaching
we are involved in the bachelor and
master programs Information Technology, where we are committed to
giving high quality lectures and lab
courses. In addition we provide optimum conditions for diploma and
master theses, student projects, and
internships. Established in March
2007, the group is lead by Prof. Mario Huemer.
The Mobile Systems Group works on
the design, modeling, and analysis of
future mobile and wireless systems.
We are engaged in research on algorithms and protocols, networking
theory, modeling and simulation aspects, and system architectures. Four
networking paradigms are currently
in the core of our activitites: selforganization, ad hoc relaying, cooperation, and mobility. Our target
application areas are telecommunications, networked embedded systems,
pervasive computing, and disaster
management. Our teaching covers
graduate courses on mobile networking and wireless communications and
undergraduate courses on electricity
and magnetism. Established in October 2005 the group is lead by Prof.
Christian Bettstetter.
The research of the Pervasive Computing Group is focused on theoretical and practical aspects of innovative
pervasive and ubiquitous computing.
Research topics include smart cameras, sensor networks, sensor fusion
and embedded signal processing architectures. Our research is conducted in close cooperation with national
and international partners both in
academia as well as in industry. Our
teaching activities cover fundamentals, technologies and applications of
pervasive computing for students of
Information Technology and Infor-
matics. Dedicated lab courses help
students to acquire practical experiences, and seminars address recent
topics of pervasive computing research. At the graduate level, we currently offer courses on fundamentals
and advanced topics on pervasive
computing, digital signal processors
and radio frequency identification
(RFID). At the undergraduate level,
we offer a lecture on digital system
design and a lab course on fundamentals of ICT. Founded in March
2007 the group is lead by Prof. Bernhard Rinner.
People
Embedded Systems and Signal Processing
Professor
Mario Huemer
Mario Huemer received the Dipl.Ing. degree in Mechatronics and the
Dr.techn. (Ph.D.) degree from the Johannes Kepler University of Linz, Austria, in 1996 and 1999, respectively. Before joining Klagenfurt University he
was with the University of Linz (19972000), Infineon Technologies Austria
(2000-2002), the University of Applied
Sciences of Upper Austria (2002-2004),
and the University of Erlangen-Nurem-
berg, Germany (2004-2007), where he
held the position of an associate professor. In March 2007 he was appointed as
a full professor for embedded systems
at Klagenfurt University. He authored
and co-authored 20 articles in journals,
magazines, and books, and about 70 papers in conference proceedings. He is
member of the IEEE, VDE, EURASIP, and ÖVE.
Office Management
Ursula Rotter
Member since 02.04.2007
Ursula Rotter was born in Klagenfurt
in 1974. After A-level at the secondary
educational establishment for financial
occupations in Klagenfurt she passed
a young entrepreneurs-training at the
WIFI. In 1995 she started her basic
training for civil service at Klagenfurt
University. During her work at the Admission Department and later at the
Faculty for Interdisciplinary Research
she started to study Education specializing in Adult and Vocational Education
which she finished in May 1995. From
1999 – 2007 she worked as secretary at
the press department. Since April 2007
she supports the Embedded Systems
and Signal Processing Group.
7
Christian Hofbauer
Research Staff member since 01.09.2007
Christian Hofbauer was born in St.
Peter am Wimberg, Austria, in 1982.
From 2002 to 2006, he studied Hardware/Software Systems Engineering at
the University of Applied Sciences of
Upper Austria. He received his Dipl.Ing. (FH) degree in 2006. He did his
internship and made his diploma thesis at the research institute IMEC in
Belgium. In September he joined the
Embedded Systems and Signal Processing Group. Currently, he focuses his research activities on physical layer aspects
of cooperative relaying.
Christian Lederer
Research and Teaching Staff member since 01.12.2007
Christian Lederer was born in Villach
in 1982. His technical education started
at the HTL for Electronics, Telecommunication and Computer Techniques in
Klagenfurt. After school he was working at the Carinthian Tech Research
(CTR) till he started his studies of Telematics at the University of Technology
in Graz. During the studies he was involved in several projects for the automotive industry and wrote his master thesis
in the field of “Chipcards” at Infineon.
In December 2007 he joined the Embedded Systems and Signal Processing
Group.
Jakob Mayring
System Administrator since 01.07.2007
Jakob Mayring was born in Munich in
1983. He moved to Klagenfurt in 2002.
During his school days he helped to
plan and deploy photovoltaic systems.
In Klagenfurt he finished the school
and enrolled in the bachelor program
of computer science at the University
of Klagenfurt in 2003. In the winter
semester 2006/2007 he worked at the
8
Information Technology Services of
the University (Zentraler Informatik
Dienst). In July 2007 he brought this
knowledge to the Embedded Systems
and Signal Processing Group where he
is working as a System Administrator.
Alexander Onic
Alexander Onic was born in 1981. He
started his studies of electrical engineering at the University of ErlangenNuremberg in 2001. After choosing
information technology as his major
subject he emphasized on signal processing and information theory in his
education. He concluded his studies and
received the Dipl.-Ing. degree in April
2007. In the same month he joined the
Embedded Systems and Signal Processing Group at Klagenfurt University. His
current research focuses on the theory
and on applications of compressive sampling.
Robert Priewasser
Robert Priewasser was born in Salzburg,
Austria, in 1982. He received the Dipl.Ing. (FH) degree at the University of
Applied Siences of Upper Austria, study
direction Hardware Software Systems
Engineering with distinction. During
his studies he spent 9 months at the
research institute IMEC in Belgium,
working on low power design for a wire-
less software defined radio platform. He
also wrote his diploma thesis at IMEC
in the field of forward error correction
codes. In autumn 2007 he started his
PhD work in Klagenfurt. The research
focus is on efficient power supplies for
embedded platforms in general, and on
PWM-based DC-DC converters specifically.
Thomas Schlechter
Thomas Schlechter was born in
Treuchtlingen, Germany, in 1979. He
received the Dipl.-Ing. degree in EEI
(Electrical Engineering, Electronics and
Information Technology) from the University of Erlangen-Nuremberg, Germany, in 2007. In April 2007 he moved
to Klagenfurt, where he is working as a
research and teaching staff member at
the Embedded Systems and Signal Processing Group. In Klagenfurt he is doing research in the field of RFID/NFC
in cooperation with the company NXP,
Gratkorn, Austria.
9
People
Mobile Systems
Professor
Christian Bettstetter studied electrical
engineering and information technology
at the Technische Universität München,
Germany, where he received the Dr.Ing. (summa cum laude) and Dipl.-Ing.
degree in 2004 and 1998, respectively.
From 2003 to 2005, Christian was a senior researcher at DoCoMo Euro-Labs.
From 1998 to 2003, he was a research
and teaching staff member at the Institute of Communication Networks
at TU München. He co-authored the
Wiley book “GSM - Switching, services
and protocols,” and published 20 articles
in journals, magazines, and books, and
more than 40 papers in conference proceedings. One of this papers received
the 2004 outstanding paper award from
the German ITG. In 2005, he was appointed as a full professor for mobile
systems at the Alpen-Adria-Universität
Klagenfurt.
Office Management
Andrea Krammer
Andrea Krammer did her language
studies at the Karl-Franzens-Universität
Graz, the Università per Stranieri and
the Università degli Studi (Italy), and
she spent a year in Milan teaching German as a foreign language. She obtained
a Master of Arts in German and Italian
(Secondary School Teacher Accreditation) (with distinction) from the Uni-
10
versity of Klagenfurt in 2001. Before
she joined the mobile systems group, she
worked as a teacher and trainer in Austria and Italy and as an office manager at
the European Academy of Sciences and
Arts in Vienna.
Helmut Adam
Helmut Adam studied telematics at TU
Graz, where he received the DI degree
in 2005. In the course of his master
thesis, which was in the field of wireless sensor networks, he worked at the
sense and control department of Infineon Technologies Austria. After his
graduation he rejoined Infineon as a
verification engineer responsible for
sensor networks and the automation of
chip verification. At the mobile systems
group, his research area is in the field of
cooperative spatial diversity in wireless
ad hoc networks. He is also tutoring an
introductory course about electrical engineering and physics for information
technology.
Günther Brandner
Günther Brandner was born in 1982. He
attended the Business Academy in Villach and worked for a Credit Institute
between September 2001 and January
2002.After military service he began
studying computer science at the University of Klagenfurt, where he received
the Master of Science with distinction
in 2007. Since 2005 he is also studying
technical mathematics. In March 2007
he joined the Mobile Systems Group as
a system administrator and junior researcher.
Sérgio Crisóstomo
Research Scholar since 01.10.2006
Sérgio Crisóstomo studied electrical
and computer engineering at the University of Porto, where he received the
MSc and Dipl.-Ing. degree in 2003 and
1997, respectively. From 1997 to 2002
he worked as a researcher at INESC
Porto. From 2002 to 2006 he was a researcher at the Laboratory of Artificial
Intelligence and Computer Science,
University of Porto, and was teaching
at the same university. In 2006, he was
granted a PhD fellowship from the Portuguese Science & Technology Foundation, and is now doing its PhD research
at the Mobile Systems Group. His topic
is network coding in dynamic networks
with focus on algorithmic and protocol
aspects.
11
Dominik Egarter
Student System Administrator since 03.12.2007
Dominik Egarter was born in 1986 in
Spittal an der Drau. He attended the
Higher Technical School in Klagenfurt
for Electrical Engineering and graduated in 2007. Since autumn 2007 he is a
student at the University of Klagenfurt.
He studies information technology. In
December 2007 he joined the Mobile
System Group as a part-time technician.
His tasks are to maintain the homepage
of the institute and to organize orders.
Furthermore he is responsible for technical support of IT infrastructure.
Wilfried Elmenreich
Senior Research Staff member since 05.11.2007
Wilfried Elmenreich graduated at the
Vienna University of Technology where
he received a Master’s degree in computer science in 1998 and a Ph.D. degree in technical sciences in 2002. His
doctoral thesis addressed the sensor fusion problem in time-triggered systems.
Wilfried contributed significantly to
the development of the TTP/A fieldbus
protocol and the standardization of the
OMG Smart Transducer Interface. In
the last five years, he has published over
40 papers in the field of embedded realtime systems. Wilfried is affiliated with
the Mobile Systems Group as a senior
researcher for Lakeside Labs, a research
center investigating self-organizing networked systems.
Michael Gyarmati
Michael Gyarmati studied telematics
at TU Graz between 1999 and 2005,
where he received the Dipl.-Ing. degree
with distinction. In parallel to telematics he has been studying for a teaching
degree in computer science and mathematics at the University of Klagenfurt
(expected graduation: 2008). In 2002 he
spent a year at the University of Leeds,
12
UK. He also worked as a software engineer at CISC Semiconductor. His
research focuses on aspects on how to
exploit mobility of mobile entities in
wireless communications and how those
findings can be adopted to benefit future
wireless applications.
Udo Schilcher
Udo Schilcher studied applied computing and technical mathematics at the
University of Klagenfurt, where he received two Dipl.-Ing. degrees (with distinction) in 2005 and 2006, respectively.
He contributed in the standardization of
the Mobile IKE protocol and the area
of number theory and cryptography. He
is now working for the mobile systems
group, where his research topic is cooperative networking and network coding. He teaches excercises in mobile and
wireless systems and a lab about network
simulation.
Alexander Tyrrell
Research Scholar since 01.11.2005
Alexander Tyrrell studied at the Ecole
Supérieure d’Ingénieurs en Electronique et Electrotechnique (ESIEE) in
Paris, where he received a master degree
in electrical engineering with a major in
signal processing and telecommunications in 2005. He also did a master of
research in digital telecommunications
systems at the Ecole Nationale Supéri-
eure des Télécommunications (ENST)
in Paris. In 2003 he spent a semester
at Chalmers University. From 2003 to
2004, he worked as a DSP support engineer at Texas Instruments in Freising,
Germany. In 2005, he started as a PhD
scholar at DoCoMo Euro-Labs in Munich, Germany, under the supervision of
Christian Bettstetter.
13
People
Pervasive Computing
Professor
Bernhard Rinner
Bernhard Rinner received both his PhD
and MSc in Telematics from Graz University of Technology in 1996 and 1993,
respectively. Before joining Klagenfurt
University he was with Graz University
of Technology and held research positions at the Department of Computer
Sciences at the University of Texas at
Austin in 1995 and 1998/99. His research
interests include parallel and distributed
processing, embedded systems as well as
mobile and pervasive computing. Bern-
hard Rinner is currently working on
pervasive computer systems, multi-DSP
architectures, embedded multimedia
systems, and distributed smart cameras.
He has authored and co-authored more
than 80 papers for journals, conferences
and workshops, lead several research
projects and served as reviewer, program
committee member, program chair and
editor-in-chief. He is member of the
IEEE and IFIP.
Office Management
Heidelies Aschbacher
Heidelies Aschbacher was born in 1977
in Klagenfurt, Austria. She attended
the Business Academy in Feldkirchen
and graduated 1996. Heidelies started
her professional education in tourism.
She was employed as front- and backoffice manager and as manager´s assistant
in hotels of premium categories and was
responsible for sales and marketing. In
14
Winter 2001 she spent 6 months working in South America. In 2005 she decided for an occupational change and
started to work as sales manager for both
national and international businesses.
Since April 2007 she supports the Pervasive Computing Group as secretary.
Wolfgang Schriebl
Wolfgang Schriebl was born in 1979
in Voitsberg, Austria. He studied Telematics at Graz University of Technology, where he did his master thesis in
the field of embedded video surveillance
systems at the Institute for Technical
Informatics (ITI). In June 2007, after working for one year as an embed-
ded software engineer at Efkon AG, he
joined the Pervasive Computing Group
at Klagenfurt University as a research
and teaching staff member. His current
research interests are in the field of pervasive smart cameras, with a focus on
smart camera architectures for distributed embedded vision.
Andreas Starzacher
Andreas Starzacher was born in 1982 in
Klagenfurt, Austria. He studied Informatics at the University of Klagenfurt
specializing in interactive systems. In
May 2007 he received his Dipl.-Ing. degree, doing his master thesis in the field
of intelligent vehicle technologies at the
Institute of Smart System Technologies. In the same month he joined the
Pervasive Computing Group as a scientific project member. He is currently
working on the “EVis” project focusing
on embedded systems technologies and
sensor fusion.
Gerald Topar
Gerald Topar was born in 1981 in Klagenfurt, Austria. He attended the HTL
in Klagenfurt for Technical Informatics and graduated in 2000. In spring
2001 he joined the company GLOCK
in Ferlach, Austria as a member of the
data processing department. In summer 2006 he received his Ing. degree.
From July 2006 till May 2007 he was
employed at the company Wernig in
Unterbergen near Ferlach, Austria.
In June 2007 he joined the Pervasive
Computing Group at the University. He
is responsible for the institute’s serverinfrastructure, the information and
communication technology laboratory,
the IT-support for the Pervasive Computing group.
15
Thomas Winkler
Thomas Winkler was born in 1980 in
Graz, Austria. He studied Telematics
at Graz University of Technology. In
2005 he earned his Dipl.-Ing. degree,
doing his thesis in the field of embedded
video surveillance systems at the Institute for Technical Informatics (ITI). In
fall 2005 he joined the Institute for Applied Information Processing and Com-
Guest Researchers
Student Assistants
• Dr. Joao Barros, Universidade de Porto, 30.31.01.2007.
• Michal Filip Gurtowski, master student in Informatics, 01.10.2007-31.01.2008
• Dr. Gerhard Bauch, TU München, 27.28.04.2007, 17.-20.04.2007.
• Francesco Mario Pivk, master student Informatics, 1.10.2007-31.01.2008
• Dr. Christian Hartmann, TU München, 15.20.05.2007
• Hima Deepthi Vankayalapati, master student Information Technology, 01.10.2007-31.01.2008
• Univ.-Prof. Dr. Joachim
München, 01.-09.06.2007.
Hagenauer,
TU
• FH-Prof. Dr. Mario Jungwirth, FH Wels, Dec.
2007-Jan. 2008.
• Dr. Ulrich Neffe, NXP Semiconductors Austria
GmbH Styria, Oct. 2007.
• Dipl.-Ing. Robert Vilzmann, TU München,
18.-24.06.2007
16
munications (IAIK) as a member of
the EU research project “Open Trusted
Computing” working on embedded security solutions. In May 2007 he joined
the Pervasive Computing Group. His
research is focused on pervasive smart
camera systems covering system architectures as well as operating system and
middleware aspects.
• Michael Strutz, bachelor student in Information
Technology, 01.10.2007-31.01.2008
• Manuel Warum, bachelor student in Informatics, 01.08.2007-21.09.2007
Research Activities and Projects
The NES institute works on the design, modeling, and
analysis of future networked and embedded systems.
Potential application areas include telecommunications, mobile computing, and disaster management.
Our project portfolio includes both long-term and
Academic Partners
Industrial Partners
• Austrian Center of
Competence in Mechatronics, Linz, Austria.
Joint research activities.
• Infineon Technologies
AG Austria. Research
contract, 01/200812/2010.
• IMEC, Belgium. Joint
master theses and internships.
• Georgia Tech, USA.
Joint research activities.
• KAI GmbH, Villach,
Austria. Two external
PhD students.
• Soongsil University,
Seoul, Korea. Research
contract.
• TU Graz, Austria.
Joint research projects
and four PhD students.
• University of ErlangenNuremberg, Germany.
Three external PhD
students.
• University of Linz,
Austria. Guest professorship
• University of Porto,
Portugal. Joint research
activities and joint PhD
student.
• Infineon Technologies
AG, Munich, Germany. Four external PhD
students
• NXP Semiconductors Austria GmbH.
Research Contract,
01/2008-12/2008.
• DICE Linz GmbH.
Joint research activities and external PhD
student.
• Siemens AG Munich,
Germany. External
PhD students and EU
project RESOLUTION.
• Austrian Research
Centers. Research contract, 11/2007-04/2008
• EFKON AG Graz,
Austria. Research project, 05/2007-04/2010
• EVK GmbH Grambach, Austria. Research
short-term research. Several projects are in close cooperation with industrial and academic partners.
The following pages give an overview of our research
activities and collaborators and explain in more detail
some selected research topics.
Main Funding
contract, 12/200702/2008.
• NXP Research Eindhoven. Joint master
theses and internships.
• The institute receives
its main funding from
the Carinthian Economic Promotion Fund
(KWF).
• France Telecom Group,
Orange Labs, France.
Research contract,
09/2006-08/2008.
Other Funding
Other Sponsors
• Portuguese Science
and Technology Fund
(FCT), Portugal, Research scholarship,
since 10/2006
• DoCoMo Communications Laboratories
Europe GmbH, Munich, Germany. Research scholarship, joint
PhD student.
• Texas Instruments
Germany GmbH. Lab
equipment.
• Middleware for network eccentric and
mobile applications
(MiNEMA), funded
by the European Science Foundation
(ESF), Steering board
member from 01/06
until 12/08
• IBM Austria. Conference Sponsorship.
• Altera Corporation,
California, USA. Lab
equipment.
• Sun Microsystems
GmbH, Vienna. Lab
equipment.
17
Embedded Systems
and Signal Processing
Overview
The Embedded Systems and Signal Processing Group
focuses on the system-, algorithmic-, architecturaland hardware-oriented aspects of embedded systems.
Application areas are short range and cellular wireless
systems, wireless positioning systems as well as automotive applications.
Our vision is to work on long term research projects as
well as on short term topics, the latter in close cooperation with industry partners.
Research projects
We succeeded in starting projects and co-operations
with excellent industry and scientific partners.
The following projects are fully or partly funded by our
research partners:
• Power DCDC: Power management for DCDC converters funded by Infineon Technologies
Austria
• RELAY: Cooperative Relaying in Wireless Networks funded by Lakeside Labs GmbH
• VHD: Very High Data Rates for Contactless
Smartcard Devices funded by NXP Semiconductors Austria GmbH
• Digital Signal Processing Concepts for RFICs
funded by LCM Linz and DICE Linz GmbH
Furthermore we work on a strategic, long term fundamental research project in the field of Compressive
Sampling/Compressive Sensing.
Project proposal submissions
The Embedded Systems and Signal Processing Group
is part of a large consortium, which submitted a largescale integrating project (IP) proposal to the European
Commission within the Objective ICT-2007.6.3 (ICT
for environmental Management and Energy Efficiency) in 2007.
Together with Infineon Technologies Austria a project
proposal for Lakeside Labs in the area of Power management for DC-DC converters is in preparation.
embedded systems and signal processing Group
Competences
18
Physical Layer Aspects of Cooperative Relaying
Christian Hofbauer and Mario Huemer
In recent years, a lot of attention has been drawn on
diversity techniques enabled by multiple antenna systems. Due to the different locations of the antennas,
spatial diversity is provided.
In combination with temporal diversity, effective
techniques have been developed in order to mitigate
the negative effects of multipath propagation, noise
or attenuation, which therefore leads to an improved
transmission. While these attributes are desirable, several problems arise when implementing these MIMO
(Multiple Input Multiple Output) concepts, especially when applying them to mobile terminals. As such,
issues like positioning several antennas on a small
handheld device become quickly an essential hurdle
for practical systems. Hence, ideas are required to still
exploit the advantages of the MIMO technique while
handling the problems of this approach in an effective
manner.
Recently, a new form of realizing spatial diversity, namely cooperative diversity, has gained significant interest in the research community. The main idea is that
several single antenna devices cooperate in some fashion in order to transfer the data from the transmitter
to the receiver. In such a network, several devices act
as relays for an active source/destination pair, leading
to spatial diversity in a distributed fashion. Thus, low
complexity transceivers can be used while utilizing the
advantages of conventional multiple antenna systems.
Furthermore, due to the low complexity, this concept
can not only be used for conventional infrastructure
networks, but it is also likely to be applicable to wireless sensor networks as well.
This project will thus perform research on cooperative diversity relaying in the context of self-organizing
wireless networks, focusing on physical layer aspects.
System related aspects in terms of how the relays will
support the transmissions, will be investigated and assessed based on their performance as well as on their
applicability for practical systems. As such, virtual antenna arrays, distributed beamforming with limited
Basic MIMO precoding model
feedback and precoding are techniques to be focused
on. Moreover, appropriate modulation schemes, such
as hierarchical modulation, will be taken into account
as well. This approach is based on the idea that the modulation scheme can be interpreted as a scheme with
high order by receivers close to the transmitter, and as
a scheme with lower order by receivers far away from
the transmitter.
Furthermore, techniques for channel estimation and
equalization will also be investigated, as they are
known to have a significant impact on the performance
and complexity of a system.
As an overall goal, the research activities aim at optimizing the bit error behaviour and/or the system’s energy efficiency. However, in order to keep the algorithms
applicable for practical scenarios, certain parameters
like the amount and kind of required channel information - from limited to full channel knowledge - and
computational complexity on one hand, and achievable
network capacity and/or link reliability on the other
hand, have to be well traded against each other.
The research methodology will mainly include simulation-based studies and implementation on off-the-shelf
hardware. From April 2008 on the project will be funded by Lakeside Labs.
19
Digital Signal Processing Concepts for RFICs
Christian Lederer and Mario Huemer
Cancelation of the TX leakage into the receive path without the need of analog SAW filters
Current transceivers for mobile communication devices
supporting UMTS and/or HSPA are operating in fullduplex mode. That means that the transmitter and the
receiver are operating at the same time, but at different
frequency bands (Frequency Division Duplex).
Transmitter signals, which exhibit much more power
compared to the receive signals, may leak into the receiver path because of non-linearities and/or out of band
emissions. These leakage signals will be processed by
the receiver path and so they disturb the signal which
should actually be received. Unfortunately the amplitudes of the leakage signals can be much higher than
the wanted signal, with the effect, that the wanted signal may be heavily distorted.
Normally excessive filtering (typically using Surface
20
Acoustic Wave (SAW) filters) is done in the RF band
to eliminate the spurious signals. But such SAW filters are discrete devices that cannot be integrated into
an IC, which furthermore increases the costs of the
whole transceiver. For higher integration, the SAW
filters have to be eliminated from the transmitter and
receiver chain. The idea is now to replace the analogue
filters by digital signal processing in the digital front
end (DFE). A further advantage of processing in the
digital domain is that digital integrated circuits scale
well with the change of the semiconductor technologies, which is not possible in the same degree for analogue integrated circuits. This research project is done
in cooperation with DICE GmbH & Co KG and with
LCM (Linz Center of Mechatronics).
Compressive Sampling / Compressed Sensing
Alexander Onic and Mario Huemer
In order to acquire analog physical signals and all contained information the analog-to-digital-conversion
needs to be designed adequately. Presently the Nyquist
frequency is the mandatory minimum sampling frequency for almost every type of signal.
In recent publications a technique called Compressive
Sampling or Compressed Sensing came up that allows sampling of some classes of signals far below the
Nyquist frequency.
The new methods allow exact signal reconstruction
from sub-Nyquist sampled data with high probability
[2]. Even higher is the probability of reconstructing the
signal within defined error bounds. These reconstruction methods lie in the field of optimization theory and
operations research.
Most signals meet the necessary preconditions for
compressive sampling and so many applications can
gain from the technique:
- Bajwa et al. [1] describe a wireless sensor network
whose signal is acquired by a fusion center using Compressive Sampling for alleviating noise and measurement deviations.
- In decentralized sensor networks power consumption is the crucial resource for system design. A lower
sampling frequency could lower power consumption
significantly.
- Analog signals can be acquired in an already compressed fashion without the use of entropy coding simply by sampling below the Nyquist frequency.
- Candes mentions medical imaging applications like
the MRT (magnetic resonance tomography) [2], in
which incomplete frequency information needs to be
dealt with on physical conditions.
- Sensing a wide spectrum used to require a very high
sampling frequency. If the signal expected to be found
only has a narrow bandwith, Compressive Sampling
can help making the process much more efficient.
- Technology Review also considers Compressed Sen
sing for image capturing devices such as cameras and
2D projection of a high-dimensional polytope, which
are of interest in sparse signal reconstruction
medical scanners as one of the 10 Emerging Technologies 2007 [3].
The ground breaking results from recent publications
in this field can have a huge impact on current `known
facts' in information theory, as well as in signal processing and (multimedia) communication technologies.
Since this topic startet evolving over the last few years
it was mostly covered on a mathematical basis. The applicational point of view from engineering side is still
underpresented and will be target for research.
References: [1] W. Bajwa, J. Haupt, A. Sayeed, and R. Nowak.
Compressive wireless sensing. In Information Processing in Sensor Networks, 2006. IPSN 2006. The Fifth International Conference on, pages 134{142, 19-21 April 2006.
[2] E.J. Candes, J. Romberg, and T. Tao. Robust uncertainty
principles: exact signal reconstruction from highly incomplete
frequency information. Information Theory, IEEE Transactions
on, 52(2):489{509, Feb. 2006.
[3] Kate Greene. Tr10: Digital imaging, reimagined. Technology
Review, http://www.technologyreview.com/Infotech/18293/,
March 2007.
21
Power DCDC - Power management for DC-DC converters
Robert Priewasser and Mario Huemer
Digital
con trol logi c
The trend in recent hardware system
designs, especially when they are battery driven (eg. notebooks, cellular
phones or handhelds), is to operate
the different components of the system (RF ciruits, baseband processor,
etc. in case of a cellular phone) with
different supply voltages, in order to
achieve optimum performance and
high power efficiency.
The task of a DC-DC converter is to
produce this variety of required supply voltages by decreasing or increasing the voltage level offered by the
source power supply (e.g. a battery).
The challenging part is to perform
this voltage conversions as power efficient as possible.
Traditionally, such DC-DC converters were implemented as analog
circuits, but recently more and more
functionality is implemented by digital circuitry. This offers more flexibility on the one hand, and offers the
opportunity to integrate more complex, maybe even adaptable control
logic, on the other hand.
The project which is done in co-operation with Infineon Villach aims to
investigate and also develop a PWMbased digital control logic for future
DC-DC converters, with the goal to
reduce the overall power consumption.
22
D
A
PW M
Ua
Ue
Schematic view of a DC-DC buck converter
with digital control logic
The Figure above shows a high
level schematic of a digitally controlled DC-DC converter is depicted.
As a first step, the analog components of the DC-DC converter
must be known in order to control
the circuit in an optimized fashion. It might also be necessary
to investigate adaptable control logic, which autonomously can estimate the parameters of the external elements (tolerances of coils,
capacitors, etc.).
With a more accurate knowledge
of the control path, the control
process can be optimized. Furthermore, power savings can
also be achieved by optimizing
the transient effects when turning the circuit on and off. This
is a common task in many modern systems, where only active
hardware blocks are powered on.
The research results shall be verified by simulations and by measurements on real prototype hardware.
Very High Data Rates for Near Field Communication (VHD)
Thomas Schlechter and Mario Huemer
Near Field Communication (NFC) is a new, short-range wireless connectivity technology that evolved from
a combination of existing contactless identification and
interconnection technologies. Products with built-in
NFC will dramatically simplify the way consumer devices interact with one another.
NFC technology is currently mainly aimed at being
used with mobile phones. Though, plenty of applications
will
be
possible
such
as:
• Mobile ticketing in public transport - an extension of the existing contactless infrastructure.
• Mobile Payment - the mobile phone acts as a
debit credit payment card.
• Smart poster - the mobile phone is used to read
RFID tags on outdoor billboards in order to get
info on the move.
• Bluetooth pairing - in the future pairing of Bluetooth 2.1 devices with NFC support will be as
easy as bringing them close together and accepting the pairing. The process of activating Bluetooth on both sides, searching, waiting, pairing
and authorization will be replaced by a simple
"touch" of the mobile phones.
• Electronic passport (containing finger prints,
photos, information for identification by the human iris, etc.)
Operating at 13.56 MHz and transferring data at up to
424 Kbits/second, NFC provides intuitive, simple, and
safe communication between electronic devices. NFC
is both a “read” and “write” technology.
Near field communication example
Communication between two NFC-compatible devices occurs when they are brought together within
about 10 centimeters range.
There are three main use cases for NFC:
• card emulation: the NFC device behaves like an
existing contactless card
• reader mode: the NFC device is active and reads
a passive RFID tag, for example for interactive
advertising
• P2P mode: two NFC devices are communicating together and exchanging information.
The underlying layers of NFC technology follow universally implemented ISO, ECMA, and ETSI standards. Because the transmission range is so short,
NFC-enabled transactions are inherently secure. Also,
physical proximity of the device to the reader gives users
the reassurance of being in control of the process.
The goal of this research project is to push the transmission rates up to several tenth of Mbits/s. The project is done in co-operation with NXP Semiconductors
Austria GmbH. Combining circuit design knowledge
in the field of RFID with modern communication
technology promises to reach the aimed high goals.
23
Mobile Systems
The research portfolio of the Mobile Systems Group is
depicted on the right side. We perform research on
networking issues in mobile and wireless systems.
This includes the design of algorithms and protocols,
contributions to networking theory, network architectures, and modelling and simulation aspects. Our goal
is to enable new applications of wireless communication that go beyond classical applications of cellular
networks and wireless local area networks.
Mobile and Wireless Networking
•
•
24
Mobility in Sparse Wireless Networks (Acronym:
SPARSE). The goal of this project is to create a
Modeling and
Simulation
Network
Architecture
Self-organization
Ad hoc networking
(multihop relaying)
Cooperative Spatial Diversity in Ad Hoc Networks. A well-known technique to improve the
robustness of wireless transmission in wireless
cellular networks is spatial diversity. In this project, we apply spatial diversity to wireless multihop networks and investigate new techniques for
packet transport based on this concept. Our main
focus is to design and assess protocols for relay
selection, packet combining, and medium access
control. The two-year project is funded by Orange
Labs, France.
Distributed Slot Synchronization in Radio Networks. We develop a distributed algorithm for slot
synchronization suited for ad hoc networks. Our
approach has been inspired from biology, from the
synchronous flashing of fireflies. We noticed that
a one-to-one transfer of the well-known firefly
synchronization to wireless networks is infeasible,
due to some characteristics of radio communications. We thus invented significant modifications,
making the synchronization converge in multihop
radio networks. Our scheme achieves high synchrony rates and a synchronization accuracy only
limited by the propagation delay. The work is done
in very close cooperation with DoCoMo EuroLabs, Munich, Germany.
Networking
Theory
Key Paradigms
Current research activities and projects are as follows:
•
Algorithms
and Protocols
Enable new applications
with focus on
Pervasive computing
Telecommunications
Exploiting cooperation
Disaster management
Exploiting mobility
Automotive
Research Portfolio Mobile Systems group
•
framework for modeling sparsely connected wireless networks with inhomogeneous node distributions and to design and asses protocols that exploit
inherent node mobility for packet delivery is such
networks. The project is partly funded by Soongsil University, Seoul, Korea within a large Korean
research project.
Network Coding. Network coding is a new, disruptive approach for transporting information
through a network. Rather than routing the information in form of dedicated packets, it enables
intermediate nodes to combine pieces of information coming from different sources. Our goal is
to investigate the impact of the network topology
on selected network coding techniques as well as
contribute to practical protocol aspects of network
coding when being applied in wireless networks.
The work is done in cooperation with the University of Porto, Portugal.
Cooperative Spatial Diversity in Ad Hoc Networks
Helmut Adam and Christian Bettstetter
The mobile radio environment is characterized by
large-scale and small-scale fading. Large-scale fading
is caused by a distance-dependent path loss and shadowing. Small-scale fading is caused by multipath propagation, i.e., radio waves are reflected and scattered at
obstacles in the environment, resulting in multiple rays
that find different ways to the receiver.
At the receiver, these multiple rays superimpose, constructively or destructively. Since the device is in general mobile and the environment changes over time due
to mobility of obstacles, these fading effects are time
variant. This leads to rapid fluctuations of the received
signal-to-noise ratio (SNR) and the delay of the different signal paths. If a mobile device moves only a small
distance, it may get from a deep “fading hole” to perfect signal reception. Indeed, the minima and maxima
of the received SNR are very close to each other and
are only separated by some fractions of the wavelength.
This is the reason why multipath fading is called smallscale fading.
The newest mitigation technique to combat the effects
of small scale fading is cooperative diversity which is
based on the concepts of spatial diversity in combination with cooperation. In traditional cellular networks, spatial diversity techniques have mainly been
investigated in a way that several antenna elements are
mounted on a single base station or device.
This is not always practically feasible on small mobile
devices. Indeed the concept of spatial diversity is very
appealing in ad hoc networks: The wireless medium
is a broadcast medium and thus devices adjacent to a
communicating device pair overhear their transmission
anyway.
Why should these devices not help to relay packets in a
cooperative manner?
The basic building block of this emerging area is the
“relay channel”: A source node transmits a message to a
destination; a third node overhears this transmission
and forwards (relays) the message to the destination;
finally, the destination combines the two received messages to improve decoding.
The main goal of this research activity is to design and
evaluate algorithms for exploiting cooperative spatial
diversity in ad hoc networks and to analyze fundamental limits of cooperative spatial diversity in multihop
environments.
Some of the results achieved so far:
•
Design and implementation of a simulation tool
to test and evaluate cooperative diversity schemas
from a protocol perspective.
•
Analysis and implementation of a schema called
“Simple Packet Combining” on wireless sensor
nodes (internship’s project) – During this project
it got evident that cooperative diversity is clearly a
cross layer task.
•
Comparison of different relay selection protocols
– the relay selection for cooperative diversity is a
key issue. So far all discovered solutions are suboptimal. The most promising one is called “Opportunistic Relay Selection”.
•
Analysis and implementation of “Opportunistic
Relay Selection” in the simulation tool.
•
Improvement of the basic “Opportunistic Relay
Selection” protocol in terms of energy efficiency
– by assessing the channel and deciding on the
reasonability of cooperation a formidable amount
of energy can be saved while keeping the overall
outage performance constant.
25
Mobility in Sparse Wireless Networks
Michael Gyarmati, Udo Schilcher, and Christian Bettstetter
The mobility of nodes is usually being experienced as
a challenge and handicap in wireless mobile networks.
Recently, however, researchers found instances in
which mobility is actually helpful for the network operation and can yield benefits. Therefore, the notion of
“exploiting mobility” gained interest in the literature.
Our intention is to investigate possibilities to exploit
the inherent mobility in wireless networks. We believe
that this would be especially useful in sparsely connected networks which can also be found in practice.
Hence, we focus on such types of networks.
Work done so far
In order to be able to accurately simulate new and existing protocols in sparse wireless networks we found
it necessary to investigate the initial spatial distribution of the nodes and how to model the mobility of
them. The initial spatial distribution defines where
on the simulation area nodes are placed. While many
researchers use the uniform distribution this is not
appropriate for sparse networks where nodes tend to
form clusters. That is why we have devised our own inhomogeneous distribution and derived its most useful
stochastic properties so that other researchers might
also use our inhomogeneous distribution.
We have also developed an objective measure to assess the inhomogeneity of any spatial distribution.
Using this measure we can automatically generate
random distributions with similar inhomogeneity as
a real location trace of a sparse network. We have also
performed an online survey to find out how the objective measure corresponds to the human perception of
inhomogeneity and how different view-angles on the
same distribution affect the human perception.
Now that we have an appropriate initial spatial distribution and allow the nodes to move it is essential
to avoid that the inhomogeneity gets lost and the distribution converges to a uniform distribution or some
26
robot for mobile test platform
other extreme case. We found that existing mobility
models cannot maintain the inhomogeneity and therefore started developing a mobility model that does. A
paper on this topic is currently in preparation and the
mobility model is being simulated and fine-tuned using a self-written mobility simulator and our inhomogeneity measure.
Outlook
After having established a profound basis for the simulation of the protocols in sparse wireless networks the
next years will be focussed on actual protocols that try
to exploit the inherent mobility. The applications we
are currently considering are mobility assisted flooding
and keeping routes stable by taking the mobility into
account when routes are established or having mobile
entities move to avoid a break-up. These protocols will
be devised, implemented, and simulated in our mobility simulator. The long-term objective is to create a real
mobile test platform based on small and cheap mobile
robots on which the protocols can be evaluated in the
real world.
Network Coding
Udo Schilcher, Sergio Crisostomo, and Christian Bettstetter
The fundamental task of communication networks is
to deliver information from source nodes to destination
nodes possibly via some intermediate nodes. This task
should be fulfilled in a reliable and resource-efficient
manner. In packet-switched networks, the information
is transported in form of packets whose payload represents the information bits that are sent by a source
node. A packet is routed throughout the communication network from the source to the destination much
in the same manner as a car moves on a road network
from the starting point to the destination point. Independent packets in the network are kept separate; routers simply store and forward the packets. This basic
assumption is common to all of today’s communication
networks, among them the worldwide Internet and
cellular wireless networks.
Recently, the concept of “network coding”, a new and
completely different paradigm for transporting information through a network was born. This concept is
based on the simple but important observation that the
act of combining different packets in a intermediate
node can yield significant advantages, leading to higher
end-to-end throughput, improvements in energy efficiency, increased robustness and security.
Our work addresses the following fields:
•
•
•
•
Visualization of network coding
Network coded flooding: comparison of established flooding techniques with network coding
Implementation and protocol aspects of network
coding in wireless systems
Existence of network codes with special properties;
size of the fields used for coding
Poster explaining Network Coding
Some of the results achieved so far are as follows:
•
•
•
•
Development of a network coding visualization
tool
Development of a simulation tool that allows experimentation and analysis of networking algorithms in dynamic networks
Analytical analysis and simulation of different approaches to flooding with and without network
coding in random graphs
Simulation results of network coding integrated
into DSR routing algorithm
27
Pervasive Computing
Pervasive Computing is the trend
towards increasingly ubiquitous,
connected computing devices in the
environment.
This trend has being leveraged by
a convergence of advanced technologies such as embedded computing, wireless communication and
sophisticated sensing. Pervasive
computing devices are not personal
computers as we tend to think of
them, but very tiny – even invisible
– devices, either mobile or embedded in almost any type of object
imaginable.
Research overview
Our research is based on the technological aspects of pervasive computing, i.e., we conduct basic and
technology-oriented research towards networked and embedded
systems. Within the comprehensive and interdisciplinary fields of
pervasive computing we focus on
the research areas sensor fusion,
embedded computing and distributed systems and demonstrate the
findings in application areas such as
traffic monitoring and intelligent
environments. Thus, our work fits
very well to the research portfolio
of the other research groups at our
institute. The work on distributed smart cameras exemplifies very
well our research activities. In this
strategic research area we focus on
28
Research Portfolio Pervasive Computing group
several aspects of distributed smart
cameras such as distributed resource management, collaborative
image processing and sensor fusion.
We develop prototypes of distributed smart camera systems and apply
them in case studies such as traffic
monitoring and security. Our research group was organizing the
“First ACM/IEEE International
Conference on Distributed Smart
Cameras” which was held in Vienna in September 2007.
Partners and Funding
Our research is conducted in close
cooperation with national and international partners both in acade-
mia as well as in industry.
National partners include Graz
University of Technology, the Austrian Research Centers, EFKON
AG Graz and the Carinthian Tech
Research, Villach. At the international level, we collaborate among
others with Georgia Institute of
Technology, Stanford University
and NXP Research. Our research
has been supported by companies
such as Texas Instruments and Altera. We are leading research projects on “Autonomous Traffic Monitoring by Embedded Vison” and
“Embedded Multi-Sensor Fusion”
– both funded by the Austrian Research Promotion Agency.
Pervasive Smart Camera Architectures for Distributed Vision
Wolfgang Schriebl and Bernhard Rinner
Pervasive Smart Cameras
Previous developments in the areas of communication
technology and embedded systems lead to smaller,
more powerful and less power-consuming system architectures. Combining these developments with the
emerging technology of smart cameras leads to smaller and cheaper intelligent camera nodes. Integrating a
large number of radio enabled cameras into a network,
makes the idea of a pervasive smart camera realizable.
The approach behind is to distribute all the smartness
of an image sensor network over a high number of sensors, to make the system acting more autonomously,
and therefore making it usable for a wider range of applications.
Vision Sensor Architecture
Designing and developing a versatile architecture for
a pervasive smart camera network in general, and an
image sensor node in particular, is rather challenging.
The capabilities of the nodes must range from distributed vision, self-calibration and self-organization to
maintenance-free operation and wireless communication. By using different types of nodes in the network,
low-level and high-level data processing can be performed on different architectures. We have proposed
an architecture for a pervasive smart camera network,
consisting of two layers of nodes with a well-defined
scope of functions.
Distributed Embedded Vision
The main power of the pervasive smart camera approach is the large number of small and cheap vision
sensors. Scene information from large areas and from
different points of view can be simultaneously grabbed.
One challenge in enabling vision on these sensors is to
cope with the large amount of image data while taking
care of the power consumption and the limited processing and communication capabilities of the embedded
system. Adapting well-known vision processes to work
Wica - wireless smart camera, prototype
in real-time, potentially on special-purpose hardware,
and to separate between in-node and in-network processing with respect to communication overhead and
information benefit, are the main goals to achieve.
Prototyping
The evaluation environment used for prototyping the
vision sensor nodes involves different technologies.
Besides general-purpose workstations in combination
with high-level languages, special-purpose hardware is
used.
Digital signal processors (DSPs) are widely used in
smart cameras for doing low-level image processing
and analysis tasks. For evaluating both, local and distributed image processing, we use DSP developer kits
provided by Texas Instruments. Currently different
approaches for object classification and object tracking
are implemented and evaluated.
Furthermore the prototype of a wireless smart camera, which combines an SIMD image processor with
a low-end general purpose processor, is provided by
NXP Research. The image processor can execute lowlevel operations on half VGA lines in a single cycle.
The special character of the parallel architecture is
currently evaluated by implementing methods for face
detection. Further work using this platform deals with
the integration of the cameras into a low-range wireless
network, and with the evaluation of algorithms for distributed object classification and object tracking.
29
On-line Embedded Multi-Sensor Data Fusion
Andreas Starzacher and Bernhard Rinner
Motivation
The world will witness a tremendous increase in the
number of vehicles in the near future. Future traffic
monitoring systems will therefore play an important
role to improve the throughput and safety of roads.
Current monitoring systems capture (vision-, acoustic, laser-based) traffic data from a large sensory network.
The overall aim of this research is to improve the quality of this traffic data by fusing the data captured from
the multiple and heterogeneous sensors.
Field of Research
Multi-sensor data fusion (MSDF) is the process of
combining homogeneous as well as heterogeneous data
coming from several different sensors. There are significant advantages over single source data as a result of
this process such as improved overall system reliability
and robustness, reduced uncertainty, extended temporal and spatial coverage, increased confidence, improved detection/classification.
This research is performed as part of the “Autonomous
Traffic Monitoring by Embedded Vision (EVis)” project which is a collaboration among Graz University of
Technology, EFKON AG and Klagenfurt University
and is funded by the FIT-IT[visual computing] program. (For more information see http://pervasive.uniklu.ac.at/evis).
Current / Future Work
The initial phase of this project has been dominated
by an intensive literature research in order to achieve
a comprehensive overview of state-of-the-art MSDF
algorithms. The focus has been particularly directed
to the fields of machine learning and embedded computing. As fusion can be performed at different levels
of abstraction, algorithms for each of these levels have
been investigated. Special focus has been laid on inference and classification methods which are promising
for resource-constrained embedded systems. An early
30
A three-layered multi-sensor data fusion
architecture
prototype of an embedded fusion architecture has been
developed. In this layered fusion architecture, each
layer is responsible for a very specific task during the
whole distributed fusion process [1]. Besides finding
novel ways of combining data exploiting spatio-temporal relations in traffic monitoring environments, the
tasks for each individual layer within the fusion architecture need to be defined precisely and implemented.
In the future we will focus on enhancing our prototype architecture evaluating the different fusion algorithms our architecture and demonstrating our approach and several real-world scenarios. We identified
tracking cars along intersections as our first case study.
Publications
[1] A. Starzacher, B. Rinner. An Embedded Multi-Sensor
Data Fusion Framework for Enhancing Vision-based Traffic Monitoring. In Proc. Intern. Conf. Distributed Smart
Cameras (PhD Forum), Vienna, 2007
Wireless Sensor Networks for Visual Computing Applications
Thomas Winkler and Bernhard Rinner
Smart Camera Networks
Smart cameras are characterized by the fact that they
provide significant amounts of computational capabilities for video and image processing and analysis. The
main benefit of this approach is that video data must
not be streamed to a centralized processing facility but
only detected events are reported to consumers of the
information. Most of the proposed camera systems rely
on fixed infrastructure in terms of power supply and
wired networking. Additionally, these systems are deployed in a static manner requiring precise calibration
by experts. With the advances in the field of wireless
sensor networks it becomes feasible to design and implement camera systems that no longer require a fixed
infrastructure. This opens up the possibility for ad-hoc
deployment of camera networks where no infrastructure is available or would be too expensive or complicated to establish.
Applications and Research Questions
Potential application scenarios are short term phenomena like crowd monitoring at sporting events taking
place at a non regular basis or at varying locations
as well as disaster management and relief operations
which can not be planned ahead of time. A central
requirement for all these application scenarios is the
capability to deploy a significant number of nodes in
an ad-hoc manner by non-expert users. This directly
leads to a number of research questions that need to
be answered when designing and implementing such
a system. Selected questions are e.g. how to localize
nodes and establish clusters and neighborhood relationships with respect to overlapping or adjacent fields
of view of the cameras. Further questions are how to
efficiently deal with the addition and removal of nodes
and, related to that, the dynamic assignment of roles to
individual nodes.
Visual Sensor Network Architecture
As a first step to approach the problem, a potential architecture for a distributed wireless vision sensor net-
System architecture for a two layer wireless
network for visual computing applications
work has been defined. The design is divided into two
layers to reduce the complexity of system deployment
and management. The first layer of nodes consists of
wireless smart cameras providing specialized image
processing functionality and low range networking capabilities. The second layer is comprised of fewer general purpose nodes which, in addition to the low range networking technology feature mid-range wireless
networking. This second layer is designed to manage
the adjacent cameras within networking range. During
deployment phase the general purpose nodes act as coordinators for localization, clustering and configuration. Once the system is operational, the general purpose
nodes provide a wireless backbone based on their midrange networking capabilities and additionally run the
execution environment for the actual applications that
rely on information delivered by the camera nodes. Important questions to be answered are how to efficiently
deploy user applications into such a network and how
to extract and aggregate relevant data from the nodes
to be delivered to the users.
The next step will be the implementation of a prototype
system based SUN Spots wireless sensor nodes and the
WiCA wireless smart camera.
31
Lakeside Labs
All professors of the institute were intensively involved into the development of Lakeside Labs—a new
research center on information and communication
technologies affiliated with the university.
32
Lakeside Labs in a Nutshell
Lakeside Labs is a new independent research center
on information and communication technologies. Research at Lakeside Labs will develop concepts, technologies and algorithms for self-organizing networked
systems.
The lab is mainly funded by the European Union,
funds from the region of Carinthia, and the state of
Austria. The total financial resources of the lab comprise ~3.5 Million EUR/year; the human resources
are 40 PY/year; both for a period of 5 years.
The lab is affiliated with the University of Klagenfurt
and should serve as an “incubator” for the new ICT
institutes and technically-oriented INF institutes of
the university, which contribute 1.5 million EUR of
the resources as in-kind contributions.
Research Area
Cell phone, Internet and wireless LAN: Information
and communication technology has made its mark
during the past decade. Scientists at Lakeside Labs
will take a further step towards the future: They are
researching into ‘self-organizing networked systems’
and their practical usage. Mobile devices, such as cell
phones and notebooks, have become our constant
companions both at work and in our leisure time.
We converse with our business partners whilst traveling, and send electronic photos home from our
holidays. We surf the Internet via fast, wireless con-
nections at work, on the campus and more and more
frequently at home. Mobile and multimedia communication technology is all around us.
However, the expectations of research and industry
from this technology are considerably higher: Not
only are cell phones and computers being interconnected, but more and more everyday objects of which
we would not immediately expect this. The latter form
an ‘Internet of things’ that is intended to be of assistance to people in their daily lives.
For instance, the wireless connection of cars can
warn of accidents and traffic jams. For such, technically and economically extremely interesting, visions
to become reality, new technologies are required that
enable a spontaneous and self-organizing networking
of devices – if possible, independent of a network infrastructure. Moreover, the networked devices interact actively with their environment in that they record
information, for example, using sensors and cameras,
and influence their environment by means of actuators. The devices are quasi embedded into their environment, where they fulfill specific tasks and form the
interface between the real and virtual world.
In this context, Lakeside Labs is concerned with information and communication technologies for ‘selforganizing networked systems’, the focal point of the
research being techniques, technologies and services
for such versatile, dynamic networks. For example,
new algorithms and protocols for energy-saving radio
transmission are being developed. “We are working on
basic concepts for the design and modeling of self-organizing technical systems,” explains Christian Bettstetter.
33
In addition to technical development, one main aspect
is human usage. Two specific application areas have
been selected in which the concepts and technologies that have been developed are to be implemented
as prototypes and tested. They comprise ‘automotive
safety’ and ‘disaster management’. In these two areas,
the efficient implementation of the new technologies
is being tested in cooperation with relevant companies
and organizations.
Cars that can look around corners
‘In the year 2000 alone, 40,000 people were killed
and 1.7 million people injured in car accidents on the
streets of the European Union.
These victims and the indirect costs involved make
these streets a dangerous and expensive transport
route’, according to Professor Kyandoghere Kyamakya, who holds the Chair in Traffic Management at
the University of Klagenfurt. Driver assistance systems, such as ABS or EPS, already help to alleviate
one of the causes of these accidents – human fatigue.
Novel applications involve both pedestrians and traffic
signs.
However, these developments have their limitations.
Just like people, they cannot see around corners and
obstacles. ‘This is where the idea of interaction and
cooperation between cars via self-organizing wireless networks is effective. Cars that can receive and
interpret information from their environment, and exchange this data with other cars in their vicinity, can
recognize and avoid dangerous situations in advance’,
says Kyamakya.
34
Technology that saves lives
The NES institute will mainly contribute to the second application area: disaster management. For various reasons, the climate change the incidence of floods,
avalanches and devastating storms is increasing. There
are also earthquakes and threats by human beings,
such as fires and terrorist attacks.
Information and communication technology plays a
central part in the warning and management of catastrophes. Thus areas and regions can be surveyed,
for example, with the aid of wireless sensor networks,
and people warned of storms, floods and avalanches in
sufficient time.
The emergency services also profit from spontaneous
wireless networking: Firefighters are being equipped
with networked cameras and displays. Earthquake relief workers communicate even when the entire network infrastructure has been destroyed.
Self-organization: a ubiquitous
phenomenon
Research into self-organizing networked systems not
only has technical and user-oriented aims, it also enables a high degree of interdisciplinary. We encounter
self-organizing systems on an almost daily basis: In
the formations of swarms of fish and migratory birds,
the interplay of termites when they build their hills, or
the activity of body cells during the healing of wounds.
In many areas of nature, single individuals or organ-
research areas and Application Areas of lakeside labs
isms work together without central coordination, but
in perfect harmony. Large areas of the economy have
already been functioning for many years according to
this paradigm.
In self-organizing systems, the instances involved form
decisions based on limited local knowledge. This leads
to a desired emergent behavior of the entire system.
Naturally self-organizing systems also possess many
characteristics that are of value in technical systems:
They are flexible and reliable, very adaptable and can
be extended at any time. It is therefore no wonder that
technical science has now discovered these and would
like to implement them wherever centrally coordinated
systems reach their limitations.
‘The economic future of the area of self-organization is
rated very positively’, says Professor Hermann Kopetz
of the Technical University of Vienna. ‘At present the
limitations of the complexity of centrally planned systems is visible in many places. Through the use of the
principles of self-organization we hope to better control
the ever-increasing complexity of large systems, and to
significantly increase their reliability.’
35
Colloquia
Hermann De Meer (Universität Passau)
Towards Modelling of Self-Organizing Systems
Self-organizing systems are typically characterized as
being decentrally self-managed based on autonomous
components. Typically, there is also a reference to
complex systems and an emergence of global structure
based on local interactions between components. Some
exchanges of information with the environment that
may even lead to an evolutionary change. While broad
consensus may be achievable about the attractiveness
of such a concept for characterizing complex systems,
there is a clear lack of formal methods and modelling
techniques as a prerequisite for a constructive usage.
It is our goal to investigate mathematical modelling
techniques for a more comprehensive description of
the processes triggering self-organization towards the
ultimate purpose of algorithmic and constructive applicability in a more technical setting.
Robert Vilzmann (TU München)
Medienzugriff mal anders: Mehrnutzerdetektion in selbstorganisierenden drahtlosen Netzen
Selbstorganisierende Netze leben von der direkten
Vernetzung mobiler Geräte, in gleichem Maße stören
sich diese Geräte jedoch auch gegenseitig. Die Vermeidung dieser Vielfachzugriffsinterferenz ist Aufgabe
der Medienzugriffskontrolle. Interferenz kann jedoch
auch durch geeignete Empfangssignalverarbeitung unterdrückt werden, insbesondere durch Mehrnutzerdetektion. In diesem Vortrag wird ein Medienzugriffs-
verfahren vorgestellt, dessen primäre Aufgabe nicht
die Vermeidung, sondern das "Management" von Interferenz ist. Ausgangspunkt sind dabei die prinzipielle
Funktionsweise der Mehrnutzerdetektion und die sich
daraus ergebenden Anforderungen für den Medienzugriff. Eine Analyse des Verfahrens berücksichtigt auch
Fälle, in denen nicht alle Geräte über ein gleichermaßen komplexe Empfangssignalverarbeitung verfügen.
Joachim Hagenauer (TU München)
Informationstheorie und Genetik
Bereits Shannon hat seine Dissertation über „Genetics“ verfasst, Genetik damals allerdings im Mendelschen Sinn. Wir sind der Überzeugung, dass Informations- und Codierungstheoretiker die aus der Technik
bekannten Methoden auf den in der DNA niedergelegten und nun weitgehend bekannten genetischen
Code anwenden sollte. Wir haben bisher Konzepte
wie Entropie, “Mutual Information” and Informationsdistanz auf die DNA-Sequenz angewandt. Mit der
Transinformation modellieren wir die Informationsübertragung von genetischen Defekten in der DNA zu
Krankheiten wie Schizophrenie. Es werden klinische
Daten verwendet um genetische Abweichungen in
den so genannten “Single Nucleotide Polymorphism
36
(SNPs)“ möglichen Krankheiten zuzuordnen. Verwendet wurden im Internet verfügbare DNA-Sequenzen
um die Shannonsche Selbst- und Transinformation
mithilfe bekannter Kompressionsalgorithmen wie
„context tree weighting“, DNA-Compress und PPM.
Eine entropiebasierte Metrik dient zur Klassifizierung und zur Erzeugung phylogenetischer Bäume von
Menschen und Säugetieren. Überaschenderweise erhielten wir mit diesem „naiven“ Ansatz die gleichen
Abstammungsbäume, wie sie Genbiologen mit anderen Methoden fanden. Die gleichen Methoden lassen
sich auch zur Klassifizierung von Texten und Sprachen
anwenden.
Publications
Journal, book, and magazine articles
Conference Papers
•
•
C. Bettstetter, M. Gyarmati, U. Schilcher: An
Inhomogeneous Spatial Node Distribution and
its Stochastic Properties. In Proc. ACM/IEEE
Intern. Symp. Modeling, Analysis, and Simulation
of Wireless and Mobile Systems (MSWiM). Chania,
Greece, Oct. 2007, pp 400-404.
•
D. Bichler, G. Stromberg, M. Huemer, M. Löw:
Key Generation Based on Acceleration Data of
Shaking Processes. In Proc. Intern. Conf. Ubiquitous Computing (UbiComp). Innsbruck, Austria,
Sept. 2007, pp. 304-317.
•
D. Bichler, G. Stromberg, M. Huemer: Innovative Key Generation Approach to Encrypt Wireless Communication in Personal Area Networks.
In Proc. IEEE Global Commun. Conf. (Globecom).
Washington, USA. Nov. 2007, 5 pages.
R. Mosshammer, F. Frank, M. Huemer: Neural
Network Based Path Detection for an FMCW
Positioning System. In Lecture Notes in Computer Science (LNCS): Computer Aided Systems
Theory – EUROCAST 2007, Springer, Berlin/
Heidelberg, Vol. 4739, pp. 928-935, 2007.
•
T. Buzid, M. Huemer, S. Reinhardt: SC/FDE
Combined with MIMO: An Improved Out of
Band Power and Performance via Tamed Frequency Modulation. In Proc. IEEE Sarnoff Symp.
Princeton, USA, April 2007.
•
R. Fasthuber, A. Papanikolaou, M. Huemer: An
Exploration of Hierarchical Design Implementation Flows. In Proc. Austrochip (in conjunction with
the 15th Austrian Workshop on Microelectronics).
Graz, Austria. Oct. 2007, pp. 135-142.
H. Witschnig, H. Stallinger, M. Huemer: A
Bandwidth Efficiency Optimized Frequency Domain Equalization Concept for Single Carrier
Transmission. In Computer Aided Systems Theory –
EUROCAST 2007, Springer, Berlin/Heidelberg,
Feb. 2007, pp. 944-951.
•
R. Gierlich, J. Hüttner, A. Dabek, M. Huemer:
Performance Analysis of FMCW Synchronization Techniques for Indoor Radiolocation. In
Proc. European Conf. Wireless Technologies (ECWT).
Munich, Germany. Oct. 2007, pp. 24-27.
•
•
•
•
•
A. Tyrrell, G. Auer, C. Bettstetter. Biologically
Inspired Synchronization for Wireless Networks.
In Advances in Biologically Inspired Information
Systems: Models, Methods, and Tools, Eds. Falko
Dressler and Iacopo Carreras, in Series: Studies
in Computational Intelligence, Springer, vol. 69,
pp. 47-62, 2007.
W. Hein, M. Huemer: Object Oriented Signal
Data Structures in VLSI Implementations of
Wireless Modems. In Lecture Notes in Computer Science (LNCS): Computer Aided Systems
Theory – EUROCAST 2007, Springer, Berlin/
Heidelberg, Vol. 4739, pp. 936-943, 2007.
M. Huemer, S. Reinhardt: A Single Carrier Frequency Domain Equalization Prototype System.
Journal of the European Microwave Association, Louvain-la-Neuve: European Microwave Association
(EUMA) Vol. 3, Issue 2, June 2007, pp 178-187.
M. Quaritsch, M. Kreuzthaler, B. Rinner, H.
Bischof, B. Strobl: Autonomous Multicamera
Tracking on embedded Smart Cameras. EURASIP Journal on Embedded Systems. 2007, 10 pages.
37
•
R. Gierlich, J. Hüttner, R. Mosshammer, M. Huemer: Local Indoor Positioning by FMCW-based
TDoA and RToF Measurements. In Proc. FH Science Day. Wels, Austria. Oct. 2007, pp. 38-43.
•
A. Gstöttner, J. Kruppa, M. Huemer: Modeling of
Dynamic Switching Currents of Digital VLSI IC
Modules and Verification by On-Chip Measurement. In Proc. Intern. Zurich Symp. Electromagnetic
Compatibility (EMC). Munich, Germany. Sept.
2007, pp 1-4.
•
•
•
•
•
38
work. In Proc. Intern. Conf. Artificial Intelligence
and Pattern Recognition (AIPR), Orlando, USA,
2007, pp 148 - 168.
•
A. Klausner, St. Erb, B. Rinner: DSP Based
Acoustic Vehicle Classification for Multi-Sensor
Real-Time Traffic Surveillance. In Proc. European
Signal Processing Conf. (EUSIPCO), Poznan, Poland. 2007, 5 pages.
•
A. Gstöttner, M. Huemer: Estimation of Current
Profiles for Large Digital VLSI Modules in Early
Design Phases. In Proc. Intern. Workshop Electromagnetic Compatibility of Integrated Circuits (EMC
Compo). Torino, Italy. Nov. 2007, pp. 87-90.
A. Klausner, C. Leistner, A. Tengg, B. Rinner:
An audio-visual sensor fusion approach for feature
based vehicle identification. In Proc. IEEE Intern.
Conf. Advanced Video and Signal based Surveillance
(AVSS), London, UK, Sept. 2007, 6 pages.
•
Wireless Modems. In Proc. Intern. Conf. Computer
Aided System Theory (EUROCAST). Feb. 2007, pp
318-320.
A. Klausner, A. Tengg, B. Rinner: Vehicle Classification on Multi-Sensor Smart Cameras using
Featureand Decision Fusion. In Proc. ACM/IEEE
Intern. Conf. Distributed Smart Cameras (ICDSC).
Vienna, Austria, Sept. 2007, pp 67-74.
•
W. Hein, J. Berkmann, M. Zimmermann, M.
Huemer: Symbol Data Organization in a WBCDMA Modem. In Proc. European Conf. Wireless
Technologies (ECWT). Munich, Germany. Oct.
2007, pp. 12-15.
R. Mosshammer, F. Frank, M. Huemer: Neural
Network Based Path Detection for an FMCW
Positioning System. In Proc. Computer Aided Systems Theory (EUROCAST). Springer, Berlin/Heidelberg. Feb. 2007, pp. 928-935.
•
M. Jovanovic, B. Rinner: Middleware for Dynamic Reconfiguration in Distributed Camera
Systems. In Proc. IEEE Intern. Workshop Intelligent Solutions in Embedded Systems (WISES). Madrid, Spain, 2007, 13 pages.
R. Mosshammer, M. Huemer, R. Szumny, K.
Kurek, J. Hüttner, R. Gierlich: A 5.8 GHz Local
Positioning and Communication System. In Proc.
IEEE International Microwave Symposium (IMS).
Hawaii, USA. June 2007, pp. 1237-1240.
•
F. Ohnhäuser, M. Huemer: Reference Generation
for A/D Converters. In Proc. IEEE Intern. Symp.
Signals, Systems and Electronics (ISSSE). Montreal,
Canada. July/Aug. 2007, pp. 355-358.
A. Klausner, A. Tengg, B. Rinner: Enhanced
Least Squares Support Vector Machines for Decision Modeling in a Multi-Sensor Fusion Frame-
•
M. Quaritsch, B. Rinner, B. Strobl: Improved
Agent-Oriented Middleware for Distributed
Smart Cameras. In Proc. ACM/IEEE Intern.
Conf.Distributed Smart Cameras (ICDSC). Vienna,
Austria, Sept. 2007, pp 297-304.
•
B. Rinner, M. Jovanovic, M. Quaritsch: Embedded Middleware on Distributed Smart Cameras.
In Proc. IEEE Intern. Conf. Acoustics, Speech and
Signal Processing (ICASSP). Hawaii, USA, June
2007, 4 pages.
•
A. Tengg, A. Klausner, B. Rinner: I-SENSE: A
Light-Weight Middleware for Embedded MultiSensor Data-Fusion. In Proc. IEEE Intern. Workshop Intelligent Solutions in Embedded Systems
(WISES), Madrid, Spain, 2007, 13 pages.
•
A. Tengg, A. Klausner, B. Rinner: An improved
genetic algorithm for task allocation in distributed
embedded systems. In Proc. Genetic and Evolutionary Conference (GECCO), London, UK, 2007, 1
page.
•
A. Tengg, A. Klausner, B. Rinner: Task Allocation in Distributed Embedded Systems by Genetic Programming. In Proc. Intern. Conf. Parallel
and Distributed Computing, Applications and Technologies (PDCAT). Adelaide, Australia, 2007, 5
pages.
•
P. Tragas, A. Kalis, C. Papadias, F. Ellinger, R.
Eickhoff, T. Ussmüller, R. Mosshammer, M.
Huemer, A. Dabek, D. Doumenis, A. Kounoudes:
RESOLUTION: Reconfigurable System for Mobile Local Communication and Positioning. In
Proc. IST Mobile & Wireless Communications Summit. Budapest, Hungary, July 2007, 5 pages.
•
H. Witschnig, M. Petit, A. Springer, M. Huemer.
Noise Reduction for SC/FDE Systems by Preand Post Processing. In Proc. European Conference
on Wireless Technologies (ECWT 2007). Munich,
Germany. Oct. 2007, pp. 28-31.
Poster of the ACM/IEEE International Conference on
Distributed Smart Cameras
39
Poster Presentations and Short Papers
•
•
•
•
•
•
•
40
Wireless Modems. In Book of Extended Abstracts
Intern. Conf. Computer Aided System Theory (EUROCAST). Las Palmas, Spain. Feb. 2007, pp.
318-320.
R. Mosshammer, F. Frank, M. Huemer, R. Szumny: Neural Network Based Path Detection for
a 5.725 GHz FMCW Positioning System. In Book
of Extended Abstracts Intern. Conf. Computer Aided
System Theory (EUROCAST). Las Palmas, Spain.
Feb. 2007, pp. 316-317.
T. Schlechter, M. Huemer: Moderne drahtlose
Übertragungsverfahren mit Entzerrung im Frequenzbereich. In HF-Report. 21. Jahrgang, Nov/
Dec. 2007, pp 40-44.
T. Schlechter: Modern wireless transmission
techniques with Frequency Domain Equalization.
Chinese-German Summer School. Shanghai, China,
Sept. 2007.
W. Schriebl: Towards Pervasive Smart Cameras.
In Proc. Intern. Conf. Distributed Smart Cameras.
Vienna, Sept. 2007.
A. Starzacher: An Embedded Multi-Sensor Data
Fusion Framework for Enhancing Vision-Based
Traffic Monitoring. In Proc. ACM/IEEE Intern.
Conf. Distributed Smart Cameras. Vienna, Austria,
Sept. 2007.
T. Winkler: Towards Pervasive Smart Cameras.
In Proc. First ACM/IEEE Intern. Conf. Distributed Smart Cameras. Vienna, Austria, Sept. 2007.
An Embedded Multi-Sensor Data Fusion Framework
for Enhancing Vision-based Traffic Monitoring
Andreas Starzacher and Bernhard Rinner (Advisor)
{andreas.starzacher, bernhard.rinner}@uni-klu.ac.at
MOTIVATION
• Achieve higher vehicle detection
and classification rate
• Less false positives
• More precise occupancy calculations
• Support for robust tracking
• Decentralized approach
FUSION ARCHITECTURE
FRf
FIS
Layer 3
PFR1
• Cooperative fusion
nd
• 2 fusion layer (fusion of partial results)
PFC1
Layer 2
• Competitive and complementary fusion
• 1st fusion layer (partial fusion in node clusters)
in1
Layer 1
Layer 3
PFRk
PFCk
in2
in1
PFCm
in1
in2
ink
RDN
RDN
RDN
RDN
SN1
SN2
SNk
SNn
Layer 2
Layer 1
• Raw data acquisition
• Data normalization
Figure 1 Three-layered fusion architecture
SENSOR NETWORK
CONTRIBUTION
• Center node (CE)
• Embedded framework for decentralized
multi-sensor data fusion
Our sensor network consists of
• Special focus on developing a new fusion
algorithm for heterogeneous data in traffic
monitoring environments
(exploiting Bayesian statistics)
(back office)
• Virtual SN (FBC)
(sending of fused data to CE)
• Sensor node (SN)
(a couple of SNs within network)
CE
coded connection
SN1
FBC
SN2
...
• Hierarchically structured
(clustering of SNs)
...
SNn
• Protocol for efficient flow of partially fused
data through the network
SNk
...
Figure 2 Heterogeneous sensor network
FURTHER INFORMATION ON
Pervasive Computing Group
Klagenfurt University, Austria
http://pervasive.uni-klu.ac.at
http://pervasive.uni-klu.ac.at/evis
Lakeside B02b
9020 Klagenfurt, Austria
[email protected]
Tel. +43 463 2700 3670
Fax +43 463 2700 3679
Poster for Phd Forum at icdsc-07
Talks
•
•
•
•
•
•
•
•
H. Aghajan, F. Berry, H. Bischof, R. Kleihorst, B.
Rinner, W. Wolf: Distributed Vision Processing
in Smart Camera Networks. Tutorial at the IEEE
Conference on Computer Vision and Pattern Recognition. Minneapolis, USA, June 2007.
C. Bettstetter: Self-Organization in Communication Networks. Invited talk at the Workshop IRAMUS, organized by INSA Lyon and INRIA. Val
Thorens, France. Jan. 2007.
C. Bettstetter: Research and Teaching at Klagenfurt’s Mobile Systems Group. Invited talk at
Soongsil University, Seoul, Korea, Feb. 2007.
C. Bettstetter: Research and Teaching at Klagenfurt’s Mobile Systems Group. Invited talk at the
Electronics and Telecommunications Research
Institute (ETRI), Mobile Telecommunication Research Division, Daejeon, Korea, Feb. 2007.
•
•
•
•
•
B. Rinner. Embedded Middleware on Distributed
Smart Cameras. Talk at the IEEE International
Conference on Acoustics, Speech, and Signal Processing
(ICASSP), Hawaii, USA, Apr. 2007.
B. Rinner. Intelligente Kameras: Vier Augen sehen
mehr als zwei. Keynote talk at Woche der Informatik, Vienna, Sept. 2007
B. Rinner: Was macht Kameras intelligent? Tag
der Forschung, University of Klagenfurt, Nov.
2007.
B. Rinner: Distributed Smart Cameras. Informatik
Kolloquium, University of Paderborn, Nov. 2007.
B. Rinner: Pervasive Computing oder wie Computer unser Alltagsleben durchdringen, Eröffnungsfeier der Fakultät für Technische Wissenschaften der Universität Klagenfurt, May 2007.
C. Bettstetter: Paradigmen zur Vernetzung der
Dinge: Über Relays, XORs und Glühwürmchen.
Inaugural lecture at the University of Klagenfurt,
Austria, Sept. 2007.
M. Huemer. The Concepts of Single Carrier
Transmission with Frequency Domain Equalization. Talk at the Intern. Conf. on Computer Aided
System Theory (EUROCAST), Las Palmas, Spain,
Feb. 2007.
M. Huemer: Embedded Systems on a Chip in der
drahtlosen Übertragungstechnik. Eröffnungsfeier
der Fakultät für Technische Wissenschaften der
Universität Klagenfurt, May 2007.
M. Huemer: Was erwarten wir von den mobilen
“Alleskönnern”? Where Shannon meets Moore.
Tag der Forschung, University of Klagenfurt, Nov.
2007.
KLAGENFURT RESEARCH DAY 2007.
DEAN Martin Hitz And NES RESEARCHER Thomas Winkler
41
Scientific Appointments
Editorial Board Member
Guest Editor
•
•
•
Christian Bettstetter is Member of the Editorial
board of "ACM Mobile Computing and
Communications Review (MC2R)".
Wilfried Elmenreich is Associate Editor of Acta
Polytechnica Hungaria, Journal of Applied
Sciences at Budapest Tech Hungary.
Conference Chairs
•
Bernhard Rinner served as general conference
co-chair of the „ACM/IEEE Conference on
Distributed Smart Cameras (ICDSC-07)“, Vienna,
Sept. 2007.
Steering Boards
•
42
Christian Bettstetter is Austrian representative and
member of the steering board of the European
Science Foundation (ESF)-funded project
"Middleware for Network Eccentric and Mobile
Applications (MiNEMA)".
•
•
•
Christian Bettstetter was guest editor of a Special
issue on 'Cooperation in Wireless Networks' in
Springer Wireless Personal Communications.
Bernhard Rinner is chief editor of a special issue on
"Distributed Smart Cameras" in the Proceedings
of the IEEE.
Bernhard Rinner is guest editor of a special issue
on "Distributed Processing in Vision Networks"
in the IEEE Journal on Selected Topics in Signal
Processing.
Bernhard Rinner is guest editor of a special issue
on "Human-Centric Applications of Distributed
Camera Networks" in the EURASIP Journal on
Image and Video Processing.
Program Committee Memberships
Christian Bettstetter:
• IEEE Intern. Conf. on Communications (ICC),
Beijing, China, May 19-23, 2008.
• European Conf. on Wireless Technology
(ECWT), Munich, Germany, Oct. 2007.
• ACM/IEEE Intern. Symposium on Modeling,
Analysis, and Simulation of Wireless and Mobile Systems (MSWiM), Chania, Crete Island,
Greece, Oct. 22-26, 2007.
• Intern. Conf. on Computer Aided System Theory (EUROCAST), Las Palmas, Spain, Feb.
2007.
• ACM SIGCOMM CoNEXT, New York, USA,
Dec. 10-13, 2007.
• IEEE Intern. Symp. Personal, Indoor, and Mobile Radio Communications, Athens, Greece,
Sept. 2007.
• Intern. Conf. on Wireless Information Networks
and Systems (WINSYS), Barcelona, Spain, July
2007.
Bernhard Rinner:
• First IEEE Workshop "From Theory To Practice in Wireless Sensor Networks," Helsinki,
Finnland, June 18, 2007.
• IEEE International Conference on Intelligent
Environments (IE), Ulm, Germany, Sept. 2007
• IEEE Vehicular Technology Conf. (VTC),
Dublin, Ireland, April 2007.
• IADIS Wireless Applications and Computing
2007 Conference, Lisbon, Portugal, June 2007
• IFIP Networking, Atlanta, USA, May 2007.
Wilfried Elmenreich:
•
Mario Huemer:
IEEE International Conference on Industrial
Informatics (INDIN), Vienna, Austria, July
2007
• Workshop on Intelligent Solutions in Embedded
Systems (WISES), Madrid, Spain, June 2007
• International Workshop on ITS for Ubiquitous
Roads (UBIROADS), Marrakech, Marocco,
June 2007
• International Symposium on Logistics and Industrial Informatics (LINDI), Wildau, Germany, Sept. 2007
• Intelligent Systems Design and Applications
(ISDA), Rio de Janeiro, Brazil, Oct. 2007
• Workshop on Intelligent Solutions in Embedded
Systems (WISES), Madrid, Spain, June 2007
43
Reviewing Activities
Christian Bettstetter served as reviewer for:
•
IEEE Journal on Selected Areas in Communications (JSAC)
•
IEEE Transactions on Mobile Computing
•
European Transactions on Telecommunications
(ETT)
•
the European Commission in the FP6 projects
ACE, ENABLE, FIREWORKS, POPEYE,
and WIP.
Wilfried Elmenreich served as reviewer for:
•
IEEE Transactions on Industrial Informatics
•
IEEE Journal of Selected Topics in Signal Processing
Michael Gyarmati served as reviewer for:
•
IEEE Transactions on Mobile Computing
Mario Huemer served as a reviewer for
•
44
Center for Innovation and Technology
(ZIT: Zentrum für Innovation und
Technologie), a subsidiary of Wiener
Wirtschaftsförderungsfonds - WWFF (Vienna
Business Agency).
•
John Wiley & Sons
•
IEEE Communication Letters
•
International Journal of Electronics and Communications (AEÜ)
•
Proceedings of the European Microwave Association (Proceedings EUMA, the Journal of the
European Microwave Association)
Bernhard Rinner served as reviewer for
•
IEEE Computer
•
EURASIP Journal on Embedded Systems
•
IEEE Transactions for Circuits and Systems
•
Machine Vision and Applications Journal
•
IEEE Journal on Selected Topics in Signal
Processing
•
ACM Transactions on Embedded Computing
Systems
PhD Examination Activities
Completed PhD Theses
•
Dr. Steffen Reinhardt: „Einträgerübertragung
für Mehrantennensysteme, Synchronisationskonzepte und expertimentelle Verifikation“,
University of Erlangen-Nuremberg, Germany.
Defense and Doctoral Examination: Erlangen,
Jan. 2007. Supervisor: Mario Huemer.
External Doctoral Students
Mario Huemer supervises the following doctoral
students at the University of Erlangen-Nuremberg,
Germany:
•
Dipl.-Ing. Tufik Buzid
•
Dipl.-Ing. (FH) Ralf Mosshammer
•
Dipl.-Ing. (FH) Andreas Gstöttner
Mario Huemer supervises the following doctoral
students in cooperation with industrial partners:
•
•
•
•
•
•
Dipl.-Ing. (FH) Daniel Bichler, Infineon Technologies AG Munich
Bernhard Rinner supervises the following doctoral
students at Graz University of Technology:
• Dipl.-Ing. Milan Jovanovic
• Dipl.-Ing. Andreas Klausner
• Dipl.-Ing. Markus Quaritsch
• Dipl.-Ing. Allan Tengg
Evaluation of doctoral theses
•
Christian Bettstetter served as a evaluator and
examiner of the doctoral thesis "performance
studies of wireless multihop networks" (Henri
Koskinen) at the Helsinki University of Technology, Finland.
•
Mario Huemer served as a evaluator and examiner of the doctoral thesis “RF-Impairments in an
HSDPA and LTE Compliant Receiver FrontEnd” (Rainer Stuhlberger) at the University of
Linz, Austria. Defense: Linz, 22.11.2007.
Dipl.-Ing. Martin Krogmann, Infineon Technologies AG Munich
Dipl.-Ing. (FH) Michael Lunglmayr, Infineon
Technologies AG Munich
Dipl.-Ing. (FH) Clemens Buchacher, Infineon
Technologies AG Munich
Dipl.-Ing. Roland Gierlich, Siemens AG Munich
Dipl.-Ing. Frank Ohnhäuser, Texas Instruments
Erlangen-Tennenlohe
45
Teaching
In the winter semester 2006/07, the Institute of Networked and Embedded Systems (NES) and the Institute of Smart System Technologies (IST) launched
a bachelor program “Informationstechnik” and a new
master program “Information Technology”.
The new programs complement the already existing
programs Informatics, Technical Mathematics, and
Information Management that have been offered by
the Department for several years.
At the time of writing this report, almost 100 students
are enrolled in one of the new programs.
Bachelor Program
The bachelor program has a duration of six semesters
with a total of 180 ECTS credits and is taught in German.
The courses in the first four semesters in the bachelor
program teach the fundamentals of electrical engineering, informatics, and engineering mathematics. The
core modules are as follows:
• Circuits and Electronics
• Signals and Systems
• Computer, Network, and Control Engineering
• Informatics and Software Engineering
• Engineering Mathematics
Each of these modules consists of compulsory lectures
and exercises. Furthermore, a comprehensive laboratory course is offered.
In the last two semesters of the undergraduate program, students have the possibility to choose courses
from application-oriented areas of information and
communication technology.
The acquired knowledge is trained in seminars, bachelor theses, and in either a team project or an internship. Last but not least, students can enroll in elective
courses to broaden one’s mind and to train soft skills.
46
Upon completion of the program, students obtain the
academic degree “Bachelor of Science (BSc)”.
Master Program
The graduate program provides the scientific qualifications necessary for a career in the development and
operation of modern information and communication
technologies. It has a duration of four semesters with a
total of 120 ECTS credits and is taught in English.
Students can choose one of six areas as their major field
of specialization. The following areas are available:
• Mobile and Wireless Systems
• Intelligent Transportation Systems
• Pervasive Computing
• Media Engineering
• Embedded Systems
• Applied Mechatronics
In addition to this technical specialization, the program also enables students to choose a methodological
specialization. The following so-called “tracks” are offered:
• Research Track
• Entrepreneurship Track
• Industrial Track
With this design concept, students have the flexibility
to concentrate on the area in which they later wish to
work.
The academic degree upon completion of the program
is Diplomingenieur/in (Dipl.-Ing.).
A description of both programs can be found at http://
tfb.uni-klu.ac.at.
The institute’s teaching activities are centered around
the bachelor and master program “Information Technology” and the master program “Informatics”.
In the year 2007, several new lectures, exercises, and
lab courses have been developed and offered for the
first time. The institute covered most technically-oriented mandatory courses in the bachelor program Information Technology and offered a variety of courses
in the master program. The current teaching portfolio
comprises the following courses as well as seminars,
research seminars, and privatissima.
Bachelor Courses offered by NES in 2007
Sem Title
1 Einführung in das Studium Informationstechnik
und aktuelle Fallstudien aus der Praxis
1 Elektrotechnische und physikalische Grundlagen
der Informationstechnik
2 Schaltungstechnik
3 Signaldarstellung und -übertragung
3 Entwurf digitaler Schaltungen
4 Digitale Signalverarbeitung
Lecture
Bettstetter
Hours Exercises/Lab
1
—
Hours
Bettstetter
2
Adam
2
Huemer
Huemer
Rinner
Huemer
2
2
2
2
Schlechter
Onic, Gyarmati
Winkler
Onic
2
2
2
Master Courses offered by NES in 2007
Title
Analog Integrated Circuit Design and Simulation
Communication Theory in Mobile Systems and Genetics
Digital Signal Processors
Embedded Communications
Embedded Microcontroller Lab
Mobile and Wireless Systems I
Mobile and Wireless Systems II
Network Simulation Lab
Pervasive Computing
RFID Topics
Signal Processing Architectures for Embedded
Applications
Lecture
Jungwirth
Hagenauer
Rinner
Huemer
—
Bettstetter
Bettstetter
—
Rinner
Neffe
Huemer
Hours
2
2
2
2
2
2
2
2
2
Exercises/Lab
—
—
Schriebl
—
Schlechter
Schilcher
—
Schilcher
—
Onic
Hours
2
2
2
2
2
47
Elektrotechnische und physikalische
Grundlagen der Informationstechnik
Christian Bettstetter, Helmut Adam
Diese Einführungsvorlesung vermittelt die wichtigsten
elektrophysikalischen Grundlagen der Informationstechnik. Dabei werden elektrische und magnetische
Felder, der elektrische Strom, Entstehung elektromagnetischer Wellen, einfache Schaltungen und zeitlich
veränderliche elektrodynamische Felder eingeführt.
Fester Bestandteil der Vorlesung sind physikalische Experimente sowie die Erläuterung von Anwendungen in
Natur und Technik. Eine zweistündige Übung vertieft
die eingeführten Inhalte durch Rechenaufgaben und
studentische Kurzvorträge.
Für Studierende des Bachelor-Studiums Informationstechnik ist diese Lehrveranstaltung ein Pflichtfach im
1. Semester. Weiterhin wird sie für Studierende der
Studienrichtungen Informatik und Technische Mathematik und Datenanalyse empfohlen.
Inhalt:
0 Fachliche Einführung und Überblick
• Elektrotechnische Begriffe im täglichen Leben
• Elektromagnetische Wechselwirkung von Ladungen
• Physikalische Grundbegriffe
• Mathematische Grundlagen
• Literaturempfehlungen
1 Elektrostatik
• Elektrische Ladung
• Kräfte zwischen Ladungen
• Elektrische Feldstärke
• Arbeit, Spannung und Potential
• Elektrische Erregung
48
Elektrisches Feld zweier ungleichnamiger
Ladungen verschiedener gröSSe
• Kontinuierliche Ladungsverteilungen
• Elektrischer Fluss und eingeschlossene Ladung
• Kapazität und Kondensatoren
• Elektrostatik in Natur und Technik
2 Elektrischer Strom
Physikalische Grundlagen:
• Stromstärke und Stromdichte
• Ladungsträgerbewegung
• Widerstand und Ohm‘sches Gesetz
• Stromquelle: Die elektrische Batterie
4 Elektromagnetische Induktion
B = B ez
Physikalische Grundlagen
• Erzeugt ein Magnetfeld elektrischen Strom?
• Gesetze zur elektromagnetischen Induktion
• Zusammenhang zwischen E- und B-Feld
Elektrotechnik
z
 
• Induktivität und Spule
• Schaltungen mit Spulen: Ein- und Ausschaltvorgang
• Gegeninduktivität
Drehmoment einer Leiterschleife im Magnetfeld
• Technische Anwendungen der Induktion
5 Elektromagnetische Wellen
Einfache Schaltungen:
• Kirchhoff‘sche Regeln
• Schaltungen mit Quellen und Widerständen
• Schaltungen mit Quellen, Widerständen und
Kondensatoren
• Zusammenhang zw. E- und B-Feld (Teil 2)
• Entstehung elektromagnetischer Wellen
• Eigenschaften elektromagnetischer Wellen
• Maxwell´sche Gleichungen
• Elektromagnetische Wellen in Natur und
Technik
3 Magnetisches Feld
• Magnete
• Magnetisches Feld und Lorentzkraft
• Magnetischer Fluss und magnetische Erregung
• Ursachen von Magnetfeldern
• Magnetfelder bewegter Ladungen
• Magnetische Materialien
• Magnetische Felder in Natur und Technik
49
Schaltungstechnik
Mario Huemer, Thomas Schlechter
Die Vorlesung schließt nahtlos an die Veranstaltung
Elektrotechnische und Physikalische Grundlagen der Informationstechnik an. Nach der Betrachtung von Netzwerken und Schaltungen bei Erregung mit sinusförmigen
Spannungen und Strömen folgt die Beschreibung von
Schaltungen mit den Methoden der Laplace-Transformation. Im Zuge dessen wird u.a. der Zusammenhang
zwischen Laplace-Bereich und der (in der Technik so
wichtigen) Denkweise im Frequenzbereich hergestellt.
Einer systematischen Betrachtung von Vierpolen folgen Transistorgrundschaltungen sowie die Möglichkeiten ihrer Arbeitspunkteinstellungen, die Analyse
mit Hilfe von Ersatzschaltbildern und die Betrachtung
ihres Frequenzverhaltens. Im nächsten Schritt werden
Operationsverstärker sowie OperationsverstärkerGrundschaltungen untersucht.
Eine Anwendung von Operationsverstärkerschaltungen stellen die analogen Filter dar, die hier nicht
nur analysiert sondern auch entworfen werden. Im abschließenden Kapitel werden Schaltungsvarianten für
AD- und DA-Wandler für unterschiedliche Anwendungsgebiete vorgestellt.
Für Studierende des Bachelor-Studiums „Informationstechnik“ ist diese Lehrveranstaltung ein Pflichtfach im 2. Semester.
Die Vorlesung wird durch einen Kurs ergänzt, in dem
die in der Vorlesung gebrachten Inhalte anhand von
Rechenbeispielen weiter vertieft und ergänzt werden,
darüber hinaus haben die Studenten die Möglichkeit ,
ihr Wissen in Laboreinheiten praktisch anzuwenden.
Testaufbau einer Schaltung mit Steckboard
Inhalt:
1.Komplexe Wechselstromrechnung (Lineare Netze im eingeschwungenen Zustand)
2.Ausgleichsvorgänge in linearen Netzen
3.Elektrische Leistung bei zeitlich veränderlichen
Strömen und Spannungen
4.Lineare Vierpole
5.Transistor (Bipolar, FET) als Verstärker
6.Grundschaltungen mit mehreren Transistoren
7.Operationsverstärker und OPV-Grundschaltungen
8.Analoge Filter: Analyse und Entwurf
9.AD- und DA-Wandler
50
Entwurf digitaler Schaltungen
Bernhard Rinner, Thomas Winkler
Digitale Schaltungen spielen eine zentrale Rolle in der
Informationstechnik. Sie stellen die Grundkomponenten in der Elektronik dar und sind daher in vielen
Geräten des Alltagslebens zu finden.
Die Lehrveranstaltung führt in die Boolsche Algebra ein und behandelt im Anschluss den Entwurf von
kombinatorischen Schaltungen beispielesweise von
Multiplexern, Addieren oder Multiplizieren. Die Optimierung der entworfenen logischen Schaltungen mit
Hilfe von Karnaugh Diagrammen stellt einen wesentlichen Bestandteil der Lehrveranstaltung dar.
Speicher-Elemente wie Latches oder Flip-Flops bilden die Grundlage für den Entwurf sequentieller
logischer Schaltungen. Nach der Einführung dieser
grundlegenden Bauelemente wird auf Basis endlicher
Automaten der Entwurf einfacher sequentieller Schaltungen veranschaulicht. Der praktischen Umsetzung
digitaler Schaltungen auf gängigen programmierbaren Logik-Technologien wird ein eigener Abschnitt
gewidmet.
Im folgenden Teil der Vorlesung werden fortgeschritten Technologien wie Register Transfers und der Aufbau von Speicherzellen besprochen. Abgeschlossen
wird die Lehrveranstaltung mit der Betrachtungen
des Aufbaus eines Mikroprozessors.
Die Vorlesung wird durch einen Kurs ergänzt, in dem
die in der Vorlesung gebrachten Inhalte anhand von
Rechenbeispielen weiter vertieft und ergänzt werden.
Darüber hinaus werden die Studierenden im Rahmen des Kurses in die Grundlagen der HardwareBeschreibungssprache VHDL eingeführt.
In Laboreinheiten haben die Studierenden die Möglichkeit dieses Wissen praktisch anzuwenden und
grundlegende kombinatorische und sequentielle
Schaltungen mit Hilfe von VHDL zu realisieren.
Entwurf und Simulation einer einfachen Schaltung
Inhalte:
1.Boolsche Algebra
2.Kombinatorische Logik
3.Schaltungsoptimierung
4.Sequentielle Logik
5.Speicherelemente
6.Registertransfer
7.Aufbau eines Mikroprozessors
8.Einführung in VHDL
51
Signaldarstellung und -übertragung
Mario Huemer, Alexander Onic, Michael Gyarmati
Die Informationstechnik beschäftigt sich mit der
Übertragung, Verarbeitung und Darstellung von Informationen. Als Träger der Informationen dienen
Signale, welche mit Hilfe von Systemen manipuliert
werden, z.B. in der Form einer Filterung oder Spektralzerlegung. In der Vorlesung Signaldarstellung und
–übertragung werden die wesentlichen Grundlagen der
Signal- und Systemtheorie sowie der Übertragungstechnik behandelt.
6.Analoge Modulationsverfahren (AM, FM,
PM)
7.Grundlagen der digitalen Übertragung im Basisband
8.Grundlagen digitaler Modulationsverfahren
(ASK, PSK, FSK, QAM, GMSK)
Die Vorlesung wird von einem Kurs begleitet. Im Kurs
werden die in der Vorlesung behandelten Themengebiete mit Hilfe von Rechenbeispielen und mit Hilfe
von Simulationsbeispielen in MATLAB/SIMULINK
begleitet.
Für Studierende des Bachelor-Studiums „Informationstechnik“ ist diese Lehrveranstaltung ein Pflichtfach
im 3. Semester. Weiterhin wird die Veranstaltung für
Studierende der Studienrichtungen “Informatik” und
“Technische Mathematik und Datenanalyse” empfohlen.
Inhalte:
1.Charakterisierung von Signalen
2.Analoge
Signale
Transformation)
(Fourier-Reihen,
Fourier
3. Analoge Systeme (Beschreibungsmethoden im
Zeit- und Bildbereich)
4.Bandpasssignale- und Systeme
5.Zufallssignale und Reaktion von Systemen auf
Zufallssignale
52
DIE sinc FUNKTION
Mobile and Wireless Systems 1 and 2
Christian Bettstetter, Udo Schilcher
Contents:
1 Introduction and Overview
• History of wireless communications
• Different kinds of mobility
• Overview and classification of wireless technologies
• Key challenges in mobile and wireless systems
2 Antennas
• Antenna types
• Production of electromagnetic waves
• Energy and power of electromagnetic waves
• Radiation intensity, directivity, and gain
• Antenna reception
3 Radio propagation and channel modeling
Scope
Wide
area
Metropolitan
area
Local
area
Personal
area
GSM
UMTS
WiMax
HSDPA
WLAN 802.11
DECT
ZigBee
The lectures Mobile and Wireless Systems 1 and 2 give a
bottom-up introduction to the area of mobile and wireless communication systems. The main goal is to give
a fundamental understanding of the principles behind
wireless transmission and networking. Current technologies, such as UMTS and IEEE 802.11, are used as
examples to explain these principles. Moreover, a whole
chapter is dedicated to ad hoc and sensor networks. The
lectures are complemented by group projects, whose results are discussed in class. A tutorial course with exercises is offered for the 1st lecture.
Bluetooth
0.1
UWB
1
10
Data rate in Mbit/s
100
1000
Classification of current Wireless technologies
• Path loss and shadowing (propagation in free
space, generalized path loss models, shadow fading)
• Multipath propagation (small-scale fading, time
and frequency spread, time- and frequency-variant behavior)
• Fading mitigation techniques (diversity, equalization, frequency hopping)
4 Coding, Modulation, and Duplexing
• Representation of signals
• Conversion from analog to digital
• Channel coding (overview, block coding, convolutional coding, coding gain, channel coding in
practice)
• Digital modulation (overview, linear modula53
tion, coherent demodulation, modulation in
practice, spead spectrum modulation)
rQ
01
• Duplexing
11
5 Multiple access and Cellular concept
00
rI
• TDMA, FDMA, CDMA, SDMA
• Radio resource planning
6 Medium Access Control (MAC) protocols
• ALOHA
• Slotted ALOHA
• CSMA
• CSMA/CA
• Performance studies
7 Wireless LAN 802.11
8 Network Architecture and Mobility Protocols
• Architecture of cellular networks (General architecture, system components in GSM and
UMTS,
• Mobility in cellular networks (Addressing and
location updating, routing to mobile users,
roaming and handover)
• Mobility in the Internet (addressing and mobility problem, autoconfiguration, device mobility
with Mobile IP, service discovery)
54
10
RECEIVED SIGNAL CONSTELLATION OF QPSK MODULATION
9 Security in Mobile Networks
• Basics
• Security in GSM and UMTS
10 Ad Hoc and Sensor Networks
• Introduction and Applications
• Routing and Relaying
• Connectivity and capacity
• Wireless sensors
11 Economic, Health, and Social Aspects
Embedded Communications
Mario Huemer
This lecture course deals with the architecture of modern wireless communication systems. All main functional blocks of a mobile phone platform including
the RF transceiver and the baseband processor are addressed.
A main focus is on the board level and on the chip level
architectures. In the “Digital Baseband Transceiver”
chapter important baseband algorithms (equalization,
channel estimation, MIMO signal processing,…),
possible implementation options, and corresponding
complexity considerations are discussed. The “Analog
Transceiver “ chapter deals with the analog signal processing tasks of a wireless device, modern transmitter and receiver architectures are discussed and future
trends on the way to Software Defined Radio architectures are presented.
-Single versus Multi Carrier Techniques
(OFDM)
-Channel Estimation
-Equalization
-MIMO Concepts
•The Analog Transceiver
-Receiver Architectures
-Transmitter Architectures
Content:
• Introductory part
-Equivalent baseband representation of passband signals
-Basic principles of digital communications
-Review of digital modulation techniques
-The mobile radio channel
• Abstract Hardware View on a Mobile Phone
Terminal Platform (Functional Blocks, Partitioning, Technology, Power)
BASEBAND CHIP FOR UMTS PHONES
• The Digital Baseband Transceiver
55
Signal Processing Architectures
for Embedded Applications
Mario Huemer
This lecture course deals with signal processing algorithms and with appropriate implementation architectures that focus on embedded applications.
As a consequence low power consumption and low chip
area is of great importance for the regarded architectures. The course starts with a repetition of important
signal processing theory and algorithms (Sampling
theorem, DFT, FFT, FIR and IIR-filters).
In the following the focus lies on implementation oriented issues like fixed point effects and architecture
options. We will start with FFT architectures followed by architectures and design issues for fixed point
FIR- and IIR-filters.
An old, but nowadays more and more important algorithm, the CORDIC algorithm, possible architectures and various applications of the CORDIC will be
discussed in the next chapter. Next interpolation and
decimation filters and possible low power implementation strategies are regarded.
Another important issue in embedded applications is
the digital signal generation. We will discuss different
architectures for digital signal generators like sine wave
generators.
Finally we will regard architectures of modern DSPs
and FPGAs for embedded applications.
-DFT and FFT: theory and implementation architectures
-Digital filter basics
• Fixed point effects in digital filtering and efficient architectures
• CORDIC-algorithm: theory, architectures and
applications
• Multirate signal processing (interpolation, decimation) : theory and low power / low area HWarchitectures
• Architectures for digital signal generators (DDS,
Polynomial approximation, IIR-implementations, CORDIC)
• Architectures for digital signal processors
• FPGAs for signal processing
• Applications
In parallel a practical course will be offered, where the
effects and architectures are studied with the help of
MATLAB/SIMULINK simulations. The course starts
with an introduction in MATLAB/SIMULINK and
will then focus on practical issues on the appropriate
topics covered in the lecture course.
Content:
• Review of signal processing basics
-Characterization of digital signals
-Sampling and reconstruction, sampling theorem
56
hw-architecture for polinomial aproximation
Digital Signal Processors
Bernhard Rinner, Wolfgang Schriebl
Digital signal processors (DSPs) can be found in many
devices such as mobile phones, PDAs and digital cameras.
Their fields of application are ubiquitous ranging from
control and measurement to audio and video processing. This lecture deals with the concepts of modern
digital signal processors. It covers main topics of micro
processor architectures, focuses on the characteristic
features of DSP and presents up-to-date processors
and development methods for DSP systems. Various
case studies are presented during this lecture.
Practical Units:
• Audio processing: This unit deals with filter
functions for amplification, reverberation, modulation and with FIR-filters (low-pass, bandpass, etc.). The functions are analysed using the
waveform generator and music as a source, and
the oscilloscope and headphones as evaluation
instruments.
• Image processing: In this unit commonly used
filter functions such as the threshold operation, median-filter and box-filters (Gauss and
Sobel) are implemented. Furthermore, images
and filters are analysed and improved by using
histogram, adaptive threshold and histogramequalization.
• Image compression: Image compressing in the
frequency domain is the topic of this unit. An
image compression chain using DCT/IDCT
and huffman trees, similar to JPEG compression is realized and evaluated.
A C6416 DSP starter kit, as used by
the students in the lab
The lecture is complemented by a lab course, which
applies the theoretical knowledge of digital signal
processors to real-world problems. The students implement, under guidance of the lecturer, various digital filters for the C6416 DSP from Texas Instruments.
As programming a DSP implicates understanding the
architecture and the concepts behind, many examples
are implemented using both assembler language and
C. The implementations are executed using developer
starter kits, and are evaluated using the debugger and
profiler. The lab is organized in units, which cover typical tasks from the areas of audio processing, image
processing and optimization.
• Optimization: In this unit, the dot-product
function is optimized step by step. Starting with
sequential code the implementation is improved
by filling delay slots, by using parallel instructions, by using world-wide optimization and by
using software pipelining. The results of each
optimization step are profiled and compared
with the results of the compiler optimizations.
Contents:
1.Introduction to micro processors
2.From micro processors to DSPs
3.Development of DSP systems
4.Examples of DSPs
5.DSP programming
6.DSP applications and case studies
57
Embedded Microcontroller Lab
Thomas Schlechter
This course gives an introduction into the field of Embedded Microcontroller Systems. The participants of
this course are expected to work on a given problem,
which can be solved by a "machine".
During the course, the "machine" will, more or less
completely, be implemented by the participants. The
very interesting part of this course is, that circuit design skills (Hardware) as well as programming skills
(C, Software) are necessary. The required knowledge
will be taught during the course (basic knowledge excluded).
In order to keep hardware complexitiy as low as possible,
a breadboard is provided to each group of the course.
On this board several components of the "machine" are
already implemented, only the electronic components
have to be soldered in and additional peripherical circuits have to be designed. At the end of the course, the
final board, built by the groups, can be taken home by
the participants.
2. Development Environment:
- Software Engineering on a Windows-System
- Producing Code, executable on the Microcontrol
ler
- Programming Device, Data Transmission to the
Microcontroller
3. Communications Interface:
- Serial Asynchronous (RS232) and Synchronous
(PS/2) Data Exchange
- Parallel bidirectional Data Exchange via a bus sy
stem
4. Communication Protocolls:
- LCD Controller HD44780, PS/2 Keyboard
- Hayes Command Set, GSM 07.07 and GSM
07.06
The software, implemented by each group, controls the
microcontroller. C is consequently used for software
implementation. Basic C-knowledge is expected from
each participant at the beginning of the course.
Proceeding in the course, the participants will learn,
how software and hardware work together, complement one another and build a complete system ("machine").
Contents:
1. Microcontroller ATmega32:
- Structure
- On-Chip-Periphery
- Time Behaviour
- Interaction of software and hardware
58
Embedded Microcontroller
Analog Integrated Circuit Design and
Simulation
Mario Jungwirth (Guest Lecture)
Most analog and mixed-signal circuit design today is
done on a CMOS bulk process, using enhancement
mode devices. This lecture will start from a basic understanding of components and circuits in analog circuit design.
A first step is to clearly delineate the difference between discrete and integrated MOSFET circuit design.
Since integrated circuit design is strongly coupled with
process technology the technology of devices used in
analog design will be presented. Components and
MOSFET amplifier configurations frequently used in
analog design will be discussed using the concept of
small-signal equivalent circuits. These stages serve as
building blocks for more complex MOSFET amplifiers and allow the introduction of MOSFET integrated
circuit design principles. Analytic calculation and numeric simulation of basic building blocks of amplifier
configurations will be discussed. Finally, a deeper insight to complex circuits like oscillators, filters, mixers
and phase-locked-loops will be given.
The importance of simulation in circuit design will be
demonstrated by simulating the operation of several
of the building block circuits using PSpice. Modifications in component parameters, loading of the input
and output will emphasize the utility of simulation
programs.
circuits design and simulation
Contents:
1.Components in analog circuit design and its
technology
2.MOSFET-Transistor
3.Basic building blocks in integrated amplifiers
4.Operational amplifier concepts
5.Signal generators / oscillators
6.Filters (SC-filters, SAW-filters, etc.)
7.Amplifiers (HF, CMOS, Bipolar, Wideband)
8.Mixers
9.PLL – Phase-Locked-Loops
10.Applications
59
Communication Theory in Mobile Systems
and Genetics
Joachim Hagenauer (Guest Lecture)
This lecture is an introductory course into the field of
communication theory.
In all modern communication systems information is
being processed in multiple stages.
In this course two of these stages are discussed. The
first one is the source coding stage which tries to minimize the information representation via optimal codes
and compression. In the second treated stage channel coding is performed which makes the transmitted
data more robust against distortions introduced by the
transfer channel.
At the beginning of this lecture the basics of information theory are revised. What is information, how to
measure it and how to represent information are some
of the answered questions.
In the source coding section the focus lies on lossless
compression. Prefix free codes like Fano and Huffman
are introduced. These codes rely on the knowledge of
the statistics of the information. As an example of a
universal source code the famous Lempel-Ziv algorithm (LZ77) is discussed. After source coding channel codes are introduced. First common expressions
and parameters are introduced. Also the capacity of
different channels is derived. Then different coding
schemes starting from the simple one bit parity code
to the modern Turbo Codes are discussed.
60
Contents:
1.Basics of Information Theory (entropy, mutual
information, Kullback distance, binary symmetric channel model)
2.Shannon/Fano Code,
3.Hoffman Code,
4.Universal Source coding (LZ77)
5.Channel Coding (Channel Capacity, Hard/
Soft Decision, Eb/N0,code rate, Shannon limit,
Hamming Distance, Interleaving, Parity Code,
Hamming Code)
6.Convolutional Code (Viterbi, Trellis, Trellis
with soft decision, Punctured Codes)
7.Iterative decoding (log liklehood ration, soft
bits, boxed-plus)
8.Turbo Codes
RFID Topics
Ulrich Neffe (Guest Lecture)
Radio Frequency Identification (RFID) Technology
has got the attention of the public after the terror attacks in 2001 and the publications of the plans for the
Future Store by the Metro group.
The terror attacks caused the usage of RFID for securing personal documents that are used to identify
persons. The future store concentrated on the identification of objects to track goods within a large retail
system. In both cases the used RFID devices are powered by an electromagnetic field but they have completely different requirements on operating distance,
data storage, performance, security or pricing.
This lecture covers the main aspects of this technology. It will start with an introduction to RFID, an
overview on the history and an outlook on this lecture.
The first section describes the different RF communication technologies dependent on the carrier frequency. The second section covers the identification of objects and persons in detail. The third section describes
the current RFID market with its privacy discussions
and gives an outlook on new applications and required
changes in technology. RFID prototyping environments are used to demonstrate the technology during
the lecture. Integrated exercises support the understanding of RFID applications.
Mifare contactless card
Contents:
1.Introduction and Overview
2.RFID Technology (Low Frequency 125kHz;
High Frequency 13.56 MHz; Ultra High Frequency ~960MHz)
3.RFID Applications (Identification of Objects,
Identification of Persons)
4.RF ID Market (Current Situation, Applications
and Market Volumes, Privacy)
5.Coming applications
6.Market volumes and pricing
7.Requirements on future technologies for RFID
61
Guest Lectures at External Institutions
Mario Huemer:
• Gastprofessur an der Johannes Kepler Universität Linz im Wintersemester 2007/08: Architekturen der Digitalen Signalverarbeitung.
• Gastvorlesung an der Universität ErlangenNürnberg im Sommersemster 2007: Architekturen der Digitalen Signalverarbeitung.
Bernhard Rinner:
• Gastvorlesung an der Technischen Universität
Graz im Sommersemester 2007: Digitale Signalprozessoren.
• Gastvorlesung an der Technischen Universität
Graz im Wintersemester 2007/08: ContextAware Computing.
62
Inaugural Lectures
Paradigmen zur Vernetzung der Dinge: Über Relays, XORs und Glühwürmchen
Professor Bettstetter hielt seine Antrittsvorlesung im September 2007. Er begann seinen
Vortrag mit einer historischen Einführung
in die Geschichte des Mobilfunks: von der
Entdeckung elektromagnetischer Wellen
bis zu heutigen 3.5G Netzen. Die Ziele des
vergangenen Jahrzehnts, so Professor Bettstetter, lagen primär auf der Steigerung der
Datenraten und Multimedia-Funktionalität.
Heute kommt die Vision der „Vernetzung der
Dinge“ hinzu, die ein Schwerpunkt seiner
eigenen Forschungstätigkeit sind. Bettstetter
erläuterte beispielhaft drei Forschungsthemen, die an seinem Lehrstuhl behandelt
werden: Kooperative Multihop Kommunikation mit „Relays“, Network Coding und
verteile Zeitsynchronisierung. Zu letztem
Thema haben sich Entwickler aus der Biologie inspirieren lassen. In Südostasien blinken
Glühwürmen völlig synchron.
Mario Huemer
“Embedded Systems on a Chip” für drahtlose Übertragungsysteme
Mario Huemer stellte seinen Lehrstuhl- und Forschungsbereich im
Rahmen der Eröffnungsfeier der
Fakultät für Technische Wissenschaften am 04. Mai 2007 vor.
Nach einem kurzen geschichtlichen
Rückblick, bei dem u.a. die österreichische Schauspielerin und Erfinderin Hedy Lamarr hervorgehoben
wird, stellt er die neuesten Trends in
Forschung und Entwicklung im Bereich des Chipdesigns - insbesondere
für Mobiltelefone - vor. Huemer unterstreicht, dass nun mit dem neuen
Informationstechnik-Studium auch
in Klagenfurt Experten in diesem
Bereich ausgebildet werden.
Bernhard Rinner
Pervasive Computing - oder wie Computer unser Alltagsleben durchdringen
Im Rahmen der Eröffnungsfeier
der Fakultät für Technische Wissenschaften am 4. Mai 2007 stellte Bernhard Rinner kurz aktuelle
Forschungsthemen des Lehrstuhls für Pervasive Computing
vor. Computer finden sich bereits
jetzt in vielen Bereichen des Alltagslebens wider – ein Trend der
insbesondere durch die rasanten
technologischen Entwicklungen
weiter anhalten wird. In naher
Zukunft werden wir von vernetzten, mobilen und „unsichtbaren“ Computersystemen umgeben sein, mit denen wir interagieren können.
63
ICDSC-07
First ACM / IEEE International Conference
on Distributed Smart Cameras
The first ACM/IEEE International Conference on Distributed Smart Cameras (ICDSC) was held in Vienna,
Austria on September 25-28, 2007. The conference attracted 117 attendees mostly from North America and
Europe who showed a strong interest in the intersection
of computer vision, embedded computing, and distributed
computing.
The conference started with a series of tutorials that introduced various topics that serve as background for distributed smart cameras. 22 technical papers were presented at the conference, covering a range of work. A series
of poster/demo sessions were also held that presented 29
posters. Some work emphasized vision algorithms, while other work emphasized distributed embedded systems.
A Ph.D. forum was also held that attracted presentations
from 12 Ph.D. candidates.
Three plenary talks were given. Feng Zhao of Microsoft
Research spoke on “Sensing platforms for World-Wide
Sensor Web.” Mubarak Shah of the University of Central
Florida spoke on “Video surveillance and monitoring using
distributed cameras.” Wilfred Philips of the University of
Ghent spoke on “Challenges for single- and multi-camera video processing.” The conference closed with a panel
session titled “Distributed smart cameras: research toys or
practical tools?”
Vienna offered a very pleasant environment for the conference. The conference itself was held at the University
of Vienna in the heart of the city. The conference dinner
included a tour of the city and of a local winery.
The second ICDSC will be held at Stanford University on
September 7-11, 2008. The conference Web site is www.
icdsc.org. We invite you to attend.
64
TI fördert IT
Die Firma Texas Instruments (TI) zeigt großes Engagement am neu entstehenden Institut für Networked
and Embedded Systems (NES) der AAU Klagenfurt.
TI strebt zusammen mit Herrn Univ.-Prof. Mario
Huemer, dem Inhaber des Lehrstuhls für Eingebettete
Systeme und Signalverarbeitung, eine tiefer gehende
Kooperation an, welche ihre Wurzeln in dessen Zeit in
Erlangen besitzt. Die Kooperation soll sich über Studentenprojekte, gemeinsame Diplomarbeiten, Dissertationen und Studentenaustauschprogramme mit TIPartneruniversitäten erstrecken.
Der Grundstein wurde in Form einer großzügigen
Spende im Rahmen des TI-Elite-Universitätsförderprogramms zur Ausrüstung eines Messlabors für Studierende und Forschungsassistenten im Bereich der an
der Universität Klagenfurt neu eingerichteten Studienrichtung Informationstechnik gelegt.
Das Labor soll den Studenten der Informationstechnik
eine praxisnahe Ausbildung ermöglichen. Entwurf und
Fertigstellung mikroelektronischer Schaltungen und
eingebetteter Systeme stehen dabei im Mittelpunkt.
Die Studierenden erlernen die theoretischen Grundlagen anschaulich und haben die Möglichkeit, aktuelle
Spitzenforschung hautnah zu erleben. Versuche für
Mobilfunkanwendungen (z.B. UMTS, HSDPA,…),
drahtlose Internetanbindung, Multimediakommunikation sowie Grundlagen der Elektrotechnik sind nur
ein kleiner Ausschnitt der sich neu bietenden Möglichkeiten.
„Absolventen mit guter elektrotechnischer Ausbildung
sind rar, TI unterstützt daher gerne die Anstrengungen
der Universität Klagenfurt hier einen neuen Schwerpunkt im Bereich der Ingenieurswissenschaften aufzubauen. Das Studium der Informationstechnik mit
seiner breiten universitären Grundlagenlehre und den
angebotenen Spezialisierungsrichtungen bietet eine
höchst interessante, fundierte, moderne und praxis-
Professoren der technischen Institute
mit Dekan Martin Hitz und Frank Ohnhäuser (TI)
nahe technische Ausbildung, die gleichzeitig eine Jobgarantie für gute Absolventen mit sich bringt.“, erklärt
Herr Ohnhäuser von TI Erlangen-Tennelohe bei seinem Besuch an der Universität Klagenfurt.
Weiterhin arbeitet TI auch an der Ausarbeitung eines
Praktikumsversuchs. Dieser wird im Sommersemester
2008 in das Lehrangebot des Lehrstuhls für Eingebettete Systeme ebenso mit einfließen, wie eine Vorlesung über Mixed-Signal Schaltungsdesign. Herr Ohnhäuser wird hierbei speziell Konzepte zum Aufbau von
Analog-Digital-Wandlern vorstellen und Problematiken beim Design aufzeigen. Die Vorlesung wird als
Blockveranstaltung angeboten werden.
Quelle: Presse&Medien | Universität Klagenfurt
65
Events
Microdrones at our Institute
Microdrones are small-scale unmanned aerial vehicles
carrying payloads such as cameras and sensors. Such
microdrones enable us to obtain a bird eye’s view of
the environment which is helpful in many applications
such as in disaster situations. We plan to use these
microdrones in a Lakeside Labs research project. In
August 2007 we had an impressive demonstration of
these devices at our institute.
Institute Retreat
From July 25–27 the Institute of Networked and Embedded Systems had its first retreat. As location hotel
Traube in Admont/Styria was chosen. Since the foundation of the institute was only back some months,
one of the main topics of this encounter was dedicated
to introduce the people and there research topics of
the different groups, as well as to present the ongoing
research topics and project partners in the different
fields.
The second point of the agenda was a workshop on the
topic “studying at NES”. Aim of this session was to
discuss how to coordinate courses and exams provided
at NES, plan advertisement activities for the Information Technology program and to find ways to motivate
more students to do their project work at NES.
66
The morning of the last day was reserved for the Professors of the three groups to present their ideas and
experiences about the topics “How to be a good researcher?”, “Steps for a good publication” and “Doing
a good Dissertation”. Finally as social event and team
building activity a rafting tour at the river Enns finished this first institute retreat.
Christmas Party 2007
This year's Institute´s Christmas party was celebrated at the
traditional Felsenkeller, near the brewery Schleppe. The
professors thanked all the staff members for their work in
the last months und highlighted the well done work. The
secretaries were appreciated with flowers.
The chair Christian Bettstetter wrote a poetry in the style
of “Lasst uns froh und munter sein”. After singing this
poetry we were surprisingly not barred of the Felsenkeller.
The song wasn't sung so bad (or loud).
Lasst uns frohe Forscher sein!
Lasst uns frohe Forscher sein
und uns in dem NES erfreun.
Lustig, lustig, tralalalala,
jetzt ist Weihnachtsfeier da,
jetzt ist Weihnachtsfeier da.
Wenn ich aufgestanden bin,
fahr ich schnell zur Arbeit hin.
alle andern sind schon da,
alle andern sind schon da.
Steht der Notebook auf dem Tisch,
hack´ ich manchmal froh und frisch.
bald ist Kaffeepause da,
bald ist Kaffeepause da.
Network Coding, Relaying,
Smart Kameras und Processing,
bald ist ein Ergebnis da,
bald ist ein Ergebnis da.
Schick´ ich schnell ein Paper raus,
Ein Reviewer schaut gut drauf.
Lustig, lustig tralalalala,
bald ist auch die Tagung da,
bald ist auch die Tagung da.
Fliegen auch die Mikrodrones,
mit an Bord die „mobile phones“.
ist die Doktorarbeit da,
ist die Doktorarbeit da.
Jetzt ist erst mal Ferienzeit,
die bringt uns viel Heiterkeit.
bald ist Weihnachtsurlaub da,
bald ist Weihnachtsurlaub da!
Lyrics by Christian Bettstetter
67
Outgoing and Incoming Students
Christian Hofbauer
Embedded Systems and Signal Processing Group
Researcher
at IMEC in Leuven, Belgium
Based on a co-operation between
the Embedded Systems and Signal
Processing group and the Belgian
research institute IMEC (Interuniversity Microelectronics Center)
in Leuven, I spent eight months
– from January 07 to August 07
– working as a researcher in the
Wireless Group of IMEC.
Encouraged by the trend, that more
and more ambitious applications
are developed for mobile devices,
which therefore will create a need
for higher transfer rates, we decided
to investigate possibilities, which
are able to support high data rate
transmissions. As such, multi antenna systems promise to handle
these requirements by transmitting
several data streams simultaneously.
However, due to the increased
number of transmissions, interferences between the streams of
one user on one hand, and between
those of other users on the other
68
Schematic mIMO Model
hand, are likely to mitigate or even
destroy the advantages inherited by
multiple antennas. Hence, strategies are required which enable high
capacity, while still keeping interferences and therefore bit error rates low. As one possible approach,
we chose to investigate precoding
techniques. More specifically, using
some knowledge about the channel,
which is fed back by the receiver,
and taking this information to preequalize the data streams already
at the transmitter, interferences
should be attenuated.
Furthermore, in order to make it
usable for a practical scenario as
well, parameters like the amount
and kind of required feedback information have to be taken into
account as well. As such, we defined our test system to be a MIMO
(Multiple Input Multiple Output)
system, where the transmitter has
knowledge about the correlation of
its antennas.
Using this information, the preco-
ding block aims at minimizing the
average minimum mean square error
among the different data streams.
In terms of error rate, simulation
results and analytical verifications
showed that an increasing correlation between the antennas only leads
to a SNR loss, but it does not change the slope of the bit error rate.
This fact is valid as long as the
channel supports the number of
independent data streams. (i.e. the
rank of the channel matrix is less or
equal to the number of independent
streams). Furthermore, the diversity order D is only defined by the
number of streams Nstr and receive
antennas Nr, but it is independent
of the number of transmit antennas
(D = Nr – Nstr +1). More in detail,
the number of transmit antennas
does only influence the coding gain
in comparison to the non-precoded
case (simple spatial multiplexing).
Natacha Castellan and Jean Aviat
Mobile Systems Group
Frankreich macht Internship in Klagenfurt
Ein Professor an unserer Heimatuniversität hat uns auf
die Arbeiten von Herrn Professor Bettstetter aufmerksam gemacht“, erzählen Natacha Castellan und Jean
Aviat von der Kontaktherstellung zur Universität Klagenfurt. Die Studierenden der Mathematik und Computerwissenschaften absolvieren ein so genanntes „Internship“ am „Institut für Vernetzte und Eingebettete
Systeme“ und sammeln wertvolle Erfahrungen für den
späteren Berufseinstieg. „Wir sind herzlich empfangen
worden und freuen uns sehr, hier mitarbeiten zu dürfen“, so der einheitliche Tenor.
Wissen durch Praxis
Im Rahmen von praxisnahen Projekten lernen die
Franzosen unter fachlicher Begleitung konkrete Planung, Umsetzung und Weiterentwicklung von technischen Anwendungen. „Ich arbeite an einem Simulator, um neue Methoden zur Nachrichtenübermittlung
von einem Mobiltelefon zum anderen zu testen“, erklärt Natacha Castellan ihren Schwerpunkt „Mobile
Netzwerke“. Jean Aviats Tätigkeit liegt im Bereich der
Netzwerkcodierung. „Ein interessantes Feld, in dem
ich auch in Frankreich arbeite.“ Die Erhöhung der
Bandbreiten, Peer-to-Peer Software und Videostreaming sind nur einige der Teilgebiete, mit denen sich
der Technologiestudent aktuell beschäftigt. Im Zuge
der Forschungen werden so die Interessen und Fähigkeiten der angehenden WissenschaftlerInnen optimal
gefördert und erweitert. „Wenn wir bei einem Problem
nicht mehr weiterkommen, bekommen wir von unseren Betreuern Udo Schilcher und Michael Gyarmati
jederzeit Hilfe“, beschreibt Natacha Castellan die hervorragende Unterstützung bei den Projekten.
Lakeside Atmosphäre
Das Internship-Programm lässt neben der intensiven
Forschungsarbeit aber auch genug Zeit, um viel von
der Atmosphäre der Umgebung mitzunehmen. „Der
Technologiepark und die Universität sind sehr schön,
professionell organisiert und die Menschen mit ihren
unterschiedlichen Nationalitäten sind alle sehr freundlich und wirklich offen für Forschungen aller Art“,
bestätigen Castellan und Aviat ihre Eindrücke. Neben
der Innenstadt von Klagenfurt haben die Studierenden
auch bereits Wien und Slowenien besucht. „Es gibt so
viel zu sehen und wir werden unseren Freunden empfehlen, unbedingt hier her zu kommen. Klagenfurt ist
der richtige Platz für anspruchsvolle Forschung“, ist
sich Jean Aviat sicher.
69
Thomas Aichholzer
Pervasive Computing Group
Internship at Siemens Corporate Research in Princeton (NJ, USA)
Thomas Aichholzer was born in 1983 in St.Veit an der
Glan (Carinthia), Austria. In June 2002 he attended the
technical school (HTL) in Klagenfurt Mössingerstraße
for technical informatics and internet engineering. In the
summer term 2003 he started to study Informatics at the
University of Klagenfurt. From April 2004 to October
2007 he worked as a web developer for the Zentralen Informtikdienst (ZID) at the University of Klagenfurt. In
October 2006 he started the bachelor study in information
technology at the same university. At the moment he is
writing his diploma thesis at Siemens Corporate Research
(SCR) in Princeton, NJ (United States).
Thomas in his office at Siemens in Princeton
I’m currently working as intern in the RFID Lab at
SRC Princeton where we are developing several solutions in the field of sensor networks. This six month
internship is part of my masters thesis on localization
topics in sensor networks. My current work deals with
the deployment of a service-oriented middleware on
an ARM architecture based embedded Linux device
called Stargate. The original middleware is supposed
to run on i86 and PC architectures, so all architecture
dependent parts have to be ported to and cross compiled
for the new target architecture. The Stargate computer
acts in the sensor network as a Gateway which routes
the data from the ZigBee (sensor) network to the usual
ethernet network so that PCs or other ethernet devices
can connect to this gateway and receive the sensor data
which they are interested in.
70
Working at SCR Princeton is an exciting experience in
a multicultural environment. The working atmosphere
is very pleasant. For example it is possible to use the
gym, to play ping pong or to go outside and play basketball during working hours. At SCR they have no
time clock to record your working hours so actually all
employees and interns have flexible time.
Technical Infrastructure
The year 2007 brought many changes within the technology infrastructure. These changes have emerged due to the growth of the institute
and the new scientific requirements regarding teaching and research.
Mainly there are three sections where these changes have taken place, a
high tech laboratory, the server infrastructure and working places for
students.
Information and communication technology laboratory
The laboratory is mainly intended for the Information Technology students. It consists of eight working places equipped with benches and
high performance workstations with high resolution 19” flat screens. The
workstations are connected through gigabit-LAN intranet und internet.
Further every working place is equipped with adjustable power supplies,
soldering stations, signal generators, oscilloscopes, circuit analyser, digital signal processing boards donated by Texas Instruments including
the Code Composer Studio and construction sets for a microcontroller
board. Additionally the workstations of the laboratory are equipped with
licenses for Quartus (Altera), Matlab and Maple.
Server infrastructure
For complex computations and intensive simulations the server infrastructure has been extended to 7 HP ProLiant blades. Each of these
blades has two Intel Xeon 3 GHz dual core processors and 10 gigabytes
of RAM. To provide sufficient disk space a fibre channel storage area
network has been integrated into the infrastructure. This results in a
total storage capacity of about 18 terabytes. To provide maximum of
scalability the phone infrastructure of the conventional telephone system
is currently being replaced by a Voice over IP solution.
Students room
To provide students a possibility to work on their topics a dedicated students room has been established. This room provides about 20 workstations which can be used by students. They are connected to the Internet
via a gigabit local area network. In addition to this a wireless access is
available all over the two buildings and is being continually expanded.
For the future there are two additional high-tech micro electronic laboratories are planned.
71
Activities at the University
Bewegter Sommer am IT-Campus
Der Sommer 2007 startete mit dem 1. IT-Ferialpraktikum. Dieses Praktikum ist eine Initiative des ITCampus Kärnten zur Förderung des IT- Nachwuchses
in Kärnten. Es ist konzipiert für SchülerInnen ab dem
15. Lebensjahr und wurde an den IT- und Mathematik-Instituten der Alpen-Adria-Universität Klagenfurt
und der Fachhochschule Kärnten abgehalten.
Weg von der Informationsflut, die die Jugendlichen
bei der Berufsorientierung eher verunsichert als unterstützt, hin zum Erleben der IT-Welt, so lautet die
Strategie des IT-Campus Kärnten. Die vielen Anmeldungen zum IT-Praktikum bestätigen diese Strategie.
Die Praktika fanden vom 9. Juli bis 7. August 2007
statt und boten vielfältige Aufgaben: Vom Testen einer neuen GPS-Software im Gelände, über Austüfteln von kniffeligen Rechenaufgaben und Erstellen
von Vorlesungsunterlagen für Erstsemestrige bis hin
zum Entwickeln dreidimensionaler Gebäudemodelle
und einer Homepage zur Vereinfachung von Verwaltungsaufgaben. Eine Gruppe baute einen Roboter
mit EM-Reife: er kann nämlich einen Ball ankicken.
Tag der Forschung 2007
Ganz unter Motto „Visual Science“ – Forschung sichtbar, begreifbar, erlebbar machen, öffnete die Klagenfurter Alpen-Adria-Universität zum zweiten Mal
ihre Pforten, und mehr als 800 Interessierte nahmen
die Gelegenheit wahr, sich über die vielfältigen Forschungsbereiche der Uni zu informieren. Forschungsinseln, Ausstellungen, Demonstrationen sowie MiniVorlesungen boten den BesucherInnen einmal mehr
unterschiedlichste Möglichkeiten, Wissenschaft von
einer ganz besonderen und oft auch neuen Seite kennen zu lernen.
72
Aber nicht nur das: Dieses Team entwickelte auch eine
didaktische Bauanleitung, die sie selbst als BetreuerInnen beim Informatik-Kurs des Uni-Talentcamps
einsetzten.
Weiters zeigten Exkursionen zum Lakeside Science
& Technology Park und Technologie-Park Villach die
Praktikums-, Diplomarbeits- und Jobmöglichkeiten in
den Technologieparks und IT-Berufsbilder auf.
Eine kompakte Studieninformation sowie eine Schnuppervorlesung gaben Einblicke in das breite informationstechnische Bildungsangebot Kärntens.
Der Sommer endete mit der 1. IT-Campus Party,
die am USI-Gelände stattgefunden hat. Ob IT-UnternehmerIn oder IT-ProfesssorIn, SchülerIn oder
IT-StudentIn, alt oder jung, weiblich oder männlich:
alle waren mit Begeisterung und Enthusiasmus beim
Wettbewerb um den schnellsten ZORB-Ball dabei.
Quelle: IT-Campus Kärnten
Forschung aktiv
Die Kulturwissenschaften stellten neben den neuesten computergestützten psychologischen Verfahren auch die Veränderungen des Gehirns während einer Psychotherapie vor. „Diese Stimulation aktiviert
die Gehirnaktivität und schlechte Musterbildungen bei den PatientInnen können sich auf diese Weise ändern“, erklärt Günther Schiepek
das komplexe System. Die Gruppe rund um die Mediensoziologin Ruth
Ayass zeigte anhand von Filmbeispielen, wie die Menschen sich Medieninhalte aneignen. Die Slawistik punktete erneut mit ihrem Fragenparcours und lockte anschließend mit einem hervorragenden Buffet aus
dem Alpen-Adria- und osteuropäischen Raum.
Die Technischen Wissenschaften stellten neben dem Aushängeschild
„Unikluni“ (der sehr aktive Roboter!) auch ihre intelligenten Kameras, die „Infogestaltungsprojekte“ im Verkehrswesen und das Projekt
„Ameise“ vor. „Die Ameise ist ein Tool, ein Simulator, der die Ausbildung von Projektmanagern verbessern soll“, beschreibt Andreas Bollin
dieses innovative Programm.
Bei den Wirtschaftswissenschaften ließen sich etliche BesucherInnen
auf ihre unternehmerischen Fähigkeiten testen und reisten anschließend
mit Google Earth um die Welt, um sich bei Veronika Gaube (IFF, SozialeÖkologie) mit www.mein-fussabdruck.at einen persönlichen ökologischen Fußabdruck vermessen zu lassen.
Info-Blitzlichter
Wissenswertes, Ergänzendes und auch völlig Neues vermittelten die
ForscherInnen mit den Mini-Vorlesungen und den „Blitzlichtern“. Mario Huemer etwa führte vor Augen, wie wichtig das
„Alleskönner“-Handy in der heutigen Zeit ist und was wir zukünftig davon erwarten können. Was die Technik von der Natur lernen kann, zeigte Wilfried Elmenreich anhand überraschender Vergleiche von Systemen dem staunenden Saalpublikum.
Rahmenprogramm
Aussteller, wie Carinthian Wood oder Carinthian Tech Research, ließen
in ihre Labors blicken. Mario Mak und Michaela Frank präsentierten
künstlerische Arbeiten, die Beatles-Ausstellung sorgte für den nötigen
Schwung sowie Bernhard Ludwig im Uni-Kino für den humorvollen
Abschluss eines rundum gelungenen Tages der Forschung 2007.
73
The Bachelor and Master programs of Information Technology were established in the winter of 2006. To inform
the public about the new possibilities in technical studies several PR activities were organized in 2007 in cooperation with IT-Campus Kärnten and the FA PRM (Partner Relation Services) of Klagenfurt University.
21.05.2007
Presentation of Information Technology at the
bilingual HAK Klagenfurt (Alexander Onic, Simone
Fuchs)
12.07.2007
Presentation of Information Technology to the
university internals (Alexander Onic)
14.09.2007
IT-Campus-Party (Alexander Onic)
30.10.2007
Presentation of Information Technology at Kaserne
Bleiburg (Alexander Onic)
74
03.11.2007
Presentation of hardware prototypes for projects
in Information Technology at the Atrio, Villach
(Thomas Schlechter)
26.11.2007
Presentation of Information Technology at Kaserne
Villach (Christian Hofbauer, Michal Gurtowski)
17.12.2007
Presentation of Information Technology at HTL
Wolfsberg (Christian Hofbauer, Michal Gurtowski)
Appointments and
Responsibilities at the University
• Head of the Institute of Networked and Embedded Systems
• Chairman of the commission for a full professorship in “Embedded Systems”
• Professor and Chair of Mobile Systems Group
• Member of the commission for a full professorship in "Measurement and Control Systems"
• Elected member of the Senate
Mario Huemer
• Professor and Chair of Embedded Systems and
Signal Processing Group
• Vice Chairman of Institute of Networked and
Embedded Systems
• Chairman of the commission for a full professorship in “ Control and Measurement Systems”
at Klagenfurt University
• Member of the curricular-commission for the
bachelor and master program “Information
Technology”
• Member of the working group „Honorary Professorship for Univ.-Prof. Dr. Michael Fischer“
at Klagenfurt University, May/June 2007
Bernhard Rinner
• Professor and Chair of Pervasive Computing
Group
• General Chair and Organizer of ICDSC-07
(First ACM/IEEE International Conference on
Distributed Smart Cameras)
• Vice Chairman of Institute of Networked and
Embedded Systems
• Member of the curricular-commission for the
bachelor and master program “Information
Technology”
• Member of the working group „Honorary Professorship for Univ.-Prof. Dr. Hans Albert“ at
Klagenfurt University, June 2007
75
Photos
76
77
78
U N I S O N Oplus
Zeitschrift der Alpen-Adria-Universität Klagenfurt
www.uni-klu.ac.at/unisono
12 der Standard
P.b.b. • Verlagspostamt 9020 • 02Z030431 M • Erscheinungsort Klagenfurt • Nr. 76/3-2007 • 20. Jg. • Preis: 1,80 Euro
Standard: Aber es gibt schon
längst synchronisierte Geräte.
Bettstetter: Ja, meistens gibt
aber ein zentrales Gerät den
Takt an. Wenn Sie Ihr Handy
einschalten, wird es von der
Basisstation des Netzanbieters
synchronisiert. Wir versuchen
die Synchronisation als selbstorganisierenden Prozess dezentral zu lösen, bei dem keiner den Takt angibt.
Standard: Warum ist das wichtig?
Bettstetter: Wenn ein zentrales Gerät ausfällt, funktioniert
nichts mehr – wir nennen das
„single point of failure“. Eine
dezentrale Synchronisierung
ist die Voraussetzung dafür,
um mobile Geräte spontan
vernetzen zu können. Solche
spontanen drahtlosen Netze
WISSEN
Eingebettet
SENIORSTUDIUM LIBERALE
Mittwoch, 11. Juli 2007
Die Zukunft der Kommunikation ist drahtlos
und vernetzt. Banal? Von wegen. Christian
Bettstetter, Informationstechnik-Professor,
arbeitet über dezentrale Synchronisierung und
Kooperation von mobilen Geräten. Hier gebe es
viel zu forschen, erklärt er Oliver Hochadel.
Standard: Wieso interessiert
sich ein Mobilfunk-Professor
für Glühwürmchen?
Bettstetter: In Asien treffen
sich Glühwürmchen zu Tausenden in Bäumen und
blinken synchron. Dies ist ein
Paarungsverhalten. Wir haben
versucht, diese Synchronisationsmethode auf drahtlose
Kommunikationsnetze zu übertragen. Aber so einfach kann
man einen Algorithmus aus
der Biologie nicht in die Technik transferieren.
LAKESIDE LABS

FORSCHUNG SPEZIAL
Wir machen es wie die Glühwürmchen
Embedded Systems (Eingebettete Systeme) sind
Rechnersysteme, die für
den Anwender weit gehend unsichtbar in Autos, Flugzeugen, Waschmaschinen, Kühlschränken oder Geräten der Unterhaltungselektronik für
bestimmte vordefinierte
Funktionen verantwortlich sind. Die Forschung
und Technologieentwicklung in diesem Umfeld
wird hierzulande spätestens seit 2002 breit gefördert. In diesem Jahr startete das Infrastrukturministerium (BMVIT) die
erste Embedded-SystemsAuschreibung des FITIT-Förderprogramms. Im
mittlerweile siebenten
Call, der noch bis 30. 7.,
12 Uhr, offen ist, werden
drei Millionen Euro vergeben. (pi)
der Standard Webtipp:
www.bmvit.gv.at
www.fit-it.at
der Standard
Forschung
Spezial
Redaktion: Bettina Stimeder (Ltg.),
Peter Illetschko (Koordination),
Klaus Taschwer
Diese Beilage entsteht mit finanzieller Unterstützung des Bundesministeriums für Wissenschaft und
Forschung, des Bundesministeriums
für Verkehr, Innovation und
Technologie, des Bundesministeriums für Wirtschaft und Arbeit,
der Österreichischen Akademie der
Wissenschaften, des Wissenschaftsfonds (FWF), der industriellen Kompetenzzentren und von EC Austria.
Die redaktionelle Verantwortung
liegt beim Standard.
http://ecaustria.at
können in vielen Bereichen
nützlich sein, zum Beispiel im
Straßenverkehr. Nehmen wir
an, es gibt einen Unfall, ein
Airbag wird gezündet, und die
Nachricht „Unfall in einem
Kilometer“ wird ohne Infrastruktur von einem Auto zum
nächsten weitergeleitet. Daran
sind die Autohersteller interessiert.
Standard: Was müssen drahtlose vernetzte Systeme noch
können?
Bettstetter: Wir arbeiten auch
am kooperativen Relaying. Ein
Gerät will an ein anderes Daten schicken; ein drittes Gerät
hört mit und hilft den beiden,
indem es die mitgehörten Daten weiterleitet. So können
mobile Geräte schneller und
energiesparender kommunizieren. Ein Sensor soll ein paar
Jahre halten, ohne dass man
die Batterie austauschen muss.
Standard: Wo liegen die Probleme?
Bettstetter: Eines unterstützt
das andere, okay – aber wenn
da zehn Geräte sind, welches
soll helfen? Hierzu braucht
man Protokolle, die bestimmen, welches Gerät die Daten
weiterleitet. Wir arbeiten
zunächst mit Simulationen;
ein wenig Informationstheorie
kommt auch noch hinzu; und
schließlich versuchen wir,
unsere Algorithmen in Prototypen zu implementieren.Normalerweise laufen Pakete hintereinander durchs Netz, jetzt
versucht man die Pakete zu
kombinieren und vermeidet so
Flaschenhälse und spart Energie. Wir arbeiten daran, wie
man dieses „Network Coding“
in drahtlosen Netzen praktisch umsetzen kann.
Standard: Ist Ihre Forschung
angewandt?
Bettstetter: Es ist eigentlich
beides: einerseits Grundlagenforschung – also: Wie funktioniert so ein verteilter Algorithmus? – und andererseits anwendungsorientiert, da unsere Ergebnisse in konkreten Anwendungen eingesetzt werden
sollen. Oft gibt es praktische
Christian Bettstetter versucht Industrieprojekte so aufzusetzen, dass sie nicht „einschnüren“. Er bevorzugt natürlich
Forschung, bei der man große Freiheiten hat und flexibel
Foto: Hochadel
agieren und reagieren kann.
Probleme, die zu ganz grundlegenden, oft mathematischen,
Fragestellungen führen. Nehmen wir etwa an, wir wollen
mit Sensoren die Luftfeuchtigkeit in einem Weinberg
messen. Jeder Sensor hat eine bestimmte Funkreichweite,
„
Manche Unternehmen
sind lästig, aber auch
bei manchen
EU-Projekten ertrinkt
man angesichts
des administrativen
Overheads.
“
z. B. zehn Meter. Wie viele
Sensoren brauche ich dann bei
einer zufälligen Verteilung?
Standard:
Grundlagenforschung ist also nicht immer der
Ausgangspunkt für das Neue?
Bettstetter: Ich sehe es iterativ
– man kann von beiden Punkten ausgehen, es ist ein Hin
und Her. Ich würde eher unterscheiden in Forschung, bei
der man große Freiheiten hat
und auch flexibel auf neueste
Ideen und Forschungsergebnisse reagieren kann, und Forschung, die stark in starre Projekte eingebettet ist.
Standard: Unternehmen denken doch an Anwendungen?
Bettstetter: Ja. Wir versuchen
Industrieprojekte aber so aufzusetzen, dass sie uns nicht
einschnüren. Unsere Partner
geben uns am Anfang die Freiheitu, es so zu machen wie
wir es für richtig halten. Den
Druck, etwas Verwertbares zu
produzieren, spüre ich nicht.
Unseren Industriepartnern ist
wichtig, dass wir gute Ergebnisse in angesehenen Zeitschriften publizieren und
Ideengeber für Patente sind.
Standard: Dann ist die Industrie weniger lästig als die
öffentliche Hand?
Bettstetter: Manche Unternehmen sind lästig, aber auch
bei manchen EU-Projekten
ertrinkt man angesichts des
administrativen Overheads.
Früher gab es noch viel mehr
freie Industrieforschung. Der
Druck, Ergebnisse zu produzieren hat zugenommen.
Dennoch kooperieren wir mit
internationalen
Topfirmen
aber auch mit lokalen Unternehmen – mit Letzteren noch
hauptsächlich in der Lehre.
Standard: Was heißt das konkret für die Lehre?
Bettstetter: Ich organisiere etwa eine Ringvorlesung, bei der
die Vortragenden aus Unternehmen vor Ort kommen und
von ihrem Berufsweg und ihrem Arbeitsalltag berichten.
Die Studierenden erhalten somit einen Eindruck, wie ihr
späteres Berufsfeld aussieht.
Standard: Wie gestaltet sich
die Arbeit an der Uni?
Bettstetter: Ich war der erste
Professor, der im Oktober 2005
in den Lakeside Park gekommen ist, in dem auch IT-Firmen angesiedelt sind. Mittlerweile sind es sechs Stiftungsprofessuren. Sie erzeugen eine
ganz andere Dynamik als ein
einzelner Neuberufener. Für
mich war das eines der Haupt-
argumente, hierherzukommen.
Die Forschungsbereiche ergänzen einander sehr gut: Es
gibt Überlappungen und Synergien, aber keine Angst unter den Kollegen, dass man
sich Themen wegnehmen
könnte. Wir koordinieren das.
ZUR PERSON
Christian Bettstetter (34)
wurde im Oktober 2005
auf den Lehrstuhl Mobile Systeme der
Universität Klagenfurt
berufen. 2007 gründete
er das Institut für
Vernetzte und Eingebettete Systeme. Von 2003
bis 2005 war er Senior
Researcher am europäischen Forschungslabor
von NTT DoCoMo in
München, dem größten
japanischen Mobilfunkanbieter. Davor
studierte Bettstetter
Elektro- und Informationstechnik an der TU
München. Seine umfangreiche Industrieerfahrung ersetzte als
äquivalente Leistung
die Habilitation und ermöglichte die frühe Berufung. (oh)
Zusammenhänge im Netz finden
Semantische Technologien stehen im Mittelpunkt eines neuen Förderprogramms
Gewisse
Wortkreationen
könnten auch User überraschen, die schon von der ITIndustrie einiges gewohnt
sind. „Information Diffusion
Across Interactive Online Media“ alias „Idiom“ ist aber
auch kein unbedingt alltägliches Projekt. Hier entwickeln
Forscher der TU Graz und der
WU Wien zusammen mit Gentics Software, Prisma Solutions und Austria.info Systems
neue Zugänge zu Internetinhalten.
„Ziel ist es, dem Benutzer
ein Werkzeug zu geben, mit
dem er Themen- und Informationsströme im Web gezielt beobachten kann“, sagt Albert
Weichselbraun vom Forschungsinstitut für Rechenintensive Methoden der WU
Wien. Über eng verbundene
Visualisierungen kann der
User etwa die Popularität von
Umweltthemen,
Produkten
oder Politparteien verfolgen.
Infos stehen durch Links zu
den Texten bereit, verwandte
Themen werden mitgeliefert.
Artikel, die im Naheverhältnis
zueinander stehen, erscheinen dazu auf geografischen
und semantischen Karten.
Außerdem kann das System
automatisch
Publikationen
identifizieren, die besonders
positiv oder negativ über
Sachverhalte berichten. Ein
Prototyp zum Thema Klimawandel ist auf www.ecore
search.net/climate zu finden.
Idiom wird mit einer halben
Million Euro vom Bundesministerium für Verkehr, Innovation
und
Technologie
(BMVIT) innerhalb des Programmes FIT-IT Semantic
Systems gefördert. „Durch explosiv wachsende Datenmengen wird der Ruf nach Lösungen zur automatischen Bewältigung immer lauter. Systeme,
die Daten eigenständig aufgrund ihrer Bedeutung klassifizieren und verarbeiten können, sind deshalb ein technologisches Muss der nächsten
Jahre“, sagt Georg Niklfeld
vom Programmabwickler FFG
(Österreichische Forschungs-
förderungsgesellschaft). Solche Zukunftsarbeit leistete
unter anderem das Projekt
„NextWrap“: Hier verfolgten
TU Wien, TU Graz sowie Lixto Software das Anliegen, die
so genannten „Wrappertechnologien“ auf die nächste Ebene zu bringen. „Wrapper“ sind
Programme, die aus relativ
unstrukturierten Webseiten
strukturierte Fakten extrahieren können. „Damit kann das
Internet wie eine große Datenbank benutzt werden. Fakten
wie Preise oder Flüge lassen
sich strukturiert in einem Datawarehouse ablegen und
dann analysieren, um beispielsweise
tagesaktuellen
Marktüberblick zu erhalten“,
erklärt Robert Baumgartner,
Lektor am Institut für Informationssysteme der TU Wien.
Ein Beispiel ist die Metasuche nach Hotels auf der Seite
der Österreich Werbung (austria.info). Daten werden von
den Homepages der jeweiligen
Portale geholt und für den
User aufbereitet. Zusätzliche
Infrastruktur betreffend Datenaustausch ist daher auch
nicht mehr notwendig. Im Gegensatz zu traditionellen
Suchmaschinen geht es bei
der Datenextraktion um Fakten in den betreffenden Sites
selbst, nicht um eine Kategorisierung oder die Relevanz von
Dokumenten.
Wissensintensiv
Dyonipos (Dynamic Ontology based Integrated Process
Optimisation) ist schließlich
eine von FIT-IT geförderte
Idee, die das Know-Center
Graz, HP Austria, TU Graz und
m2n – consulting und development umsetzen. „Es geht
um Unterstützung wissensintensiver Geschäftsprozesse in
Firmen durch semantische
Technologien“, sagt Michael
Granitzer, Division Manager
des
Know-Center
Graz.
„Durch Analyse der Eingaben
kann erkannt werden, ob der
Anwender gerade ein Dokument bearbeitet oder sich im
Internet zu einem Thema
schlau macht. Dyonipos sucht
im Unternehmen dann automatisch nach ergänzender Information. Durch die Analyse
der Tätigkeiten einzelner Mitarbeiter können außerdem
Engstellen im Betrieb betreffend den Wissensfluss festgestellt und behoben werden.“
Jetzt steht der vierte Call in
Sachen Semantic Systems offen. Das Fördervolumen beträgt dabei rund drei Millionen Euro (vorbehaltlich der
Freigabe des Förderbudgets
und der förderrechtlichen
Grundlagen). Geplante Deadline für Einreichungen ist der
23. Juli, 12.00 Uhr. Entscheidungen werden im September
mitgeteilt. Niklfeld: „Diesmal
gibt es zusätzlich einen Impuls in Richtung Lösungen,
die mit einem Blick auf spätere Volumenmärkte und Anwendungen in Großsystemen
gestaltet sind. Irgendwann
muss jede neue InformationsTechnologie auch den Schritt
in die kommerzielle Verwertung schaffen.“ (pren)
79
The Institute of Networked and Embedded Systems
(NES) of the University of Klagenfurt, Austria,
works on the design, modeling, and analysis of future
networked and embedded systems. We are engaged in
research on algorithms and protocols, architectures,
networking theory, signal processing, and hardwareoriented issues, with a strong focus on wireless and
mobile networks of embedded, pervasive devices and
sensors. Current interests include self-organization in
networks, networked cameras, cooperative wireless
communications, localization and synchronization,
and sensor fusion. Our project portfolio includes longterm and short-term research, both in close cooperation
with industrial and academic partners. Potential
application areas include telecommunications, mobile
computing, automotive, and disaster management.
Our teaching activities cover bachelor, master, and
PhD courses for the study programs “information
technology” and “informatics”. We are committed
to giving high-quality lectures, seminars, labs in the
areas of signals and systems, circuits and electronics,
mobile and wireless systems, pervasive computing,
and embedded systems.
Established in the Lakeside Science & Technology
Park in January 2007, the institute consists of three
research groups: mobile systems (Prof. Bettstetter),
embedded systems and signal processing (Prof.
Huemer), and pervasive computing (Prof. Rinner). It
currently employs 27 people.
80

Mobile Systems: Annual Report 2006

Transcription

Similar documents

Kapitel 6: Betriebssysteme - Security Engineering Group | TU

2011 - Universitätsklinikum Bonn

2006 - TZI

FLEX-6000 - Funktechnik Dathe

Bewegt ÄLTER werden mit JUDO - Dachverband für Budotechniken