lar bionics who are specializing in the fields of
Bionic Interfaces and Bio-nano measurement
devices.
Parts of the course material are included in the
following books:
A. Molisch, Wireless communications,
Wiley-IEEE Press, 2005.
A. Goldsmith, Wireless communica-
tions, Cambridge university press,
2005.
C.S.R. Murthy and B.S. Manoj, Ad Hoc
wireless networks: architectures and
protocols, Prentice Hall PTR Upper
Saddle River, NJ, USA, 2004.
E.H. Callaway, Wireless sensor net-
works: architectures and protocols,
CRC press, 2004.
D. Gislason, Zigbee Wireless Network-
ing, Newnes, 2008.
G.Z. Yang and M. Yacoub, Body sen-
sor networks, Springer-Verlag New
York Inc, 2006.
The references listed above provide a wide
range of basis knowledge. But the real chal-
lenge of the course is to present a comprehen-
sive foundation which addresses all the engi-
neering challenges and list the corresponding
practical solutions. This must be done on such
a mathematical platform which takes into ac-
count only those skills which have already
been obtained in the studies preceding the
course. The books above cannot fulfill these
objectives by themselves.
Dr. András Oláh can successfully guide the
students through the course. Furthermore, as
his research area is wireless communication,
he can expose the students to the newest re-
sults of the field.
One of the research areas of Prof. Dr. János
Levendovszky is also wireless communication
and their technological challenges (wireless
detection, channel equalization, energy aware
routing in wireless sensor networks). He lead
several projects in the filed as principal inves-
tigator and supervised numerous PhD students
as well.
As a result, the expertise and skills to pro-
vide a high level course in the field is availa-
ble.
III.
R
ESULTS
The course material has been split into 12
parts, as follows:
1. Overview of wireless communications
2. Fundamentals and technical challenges
of wireless communications
3. Wireless channel characterization and
models
4. Digital modulation
5. Detection and channel equalization
6. Multiple channel access
7. Routing protocols
8. Standardized wireless systems
9. Communication protocols for wireless
sensor networks
10. Localization algorithms and strategies
for wireless sensor networks
11. Applications of ad hoc and sensor net-
works
12. Future of wireless technology and re-
search
The approximatively 500 slides had not
been uniformly divided and dedicated to the
12 chapters, but based on the importance of
the topics.
The first half of the material deals with the
foundations of wireless communication which
is then followed by the particular challenges of
ad hoc and sensor networks from Chapter 8.
It is important to note that during the course
the students also choose a project the goals of
which can then be completed on the infrastruc-
ture available at the WSN laboratory (software
radio, Texas nodes, Xbow Mica2 mote-ok,
Android based mobile telephones, Arduino
devices). In this way, the students can grow
familiar with the standard devices and devel-
opment kits and become capable of R&D in
41
40
the field of wireless communication technolo-
gies.
IV.
S
UMMARY
The course can provide comprehensive
foundations combined with up-to-the-minute
technological details in the domain of wireless
communication technologies to the students. It
also provides the knowledge which is needed
for the engineers to keep track of the ever
changing and evolving technologies and appli-
cations.
41
An Integrated Approach to Linear- and
Nonlinear Signal Processing
Digital- and Neural Based Signal Processing & Kiloprocessor Arrays
András Oláh, Dávid Tisza, Gergely Treplán,
Kálmán Tornai
Péter Pázmány Catholic University
Faculty of Information Technology
Budapest, Hungary
[olah,tisda,trege,kami]@itk.ppke.hu
János Levendovszky
Budapest University of Technology and Economics
Department of Telecommunications
Péter Pázmány Catholic University
Faculty of Information Technology
Budapest, Hungary
levendov@hit.bme.hu
Summary— The course objective is to give a
comprehensive introduction to digital signal
processing using the traditional linear- and
nonlinear approaches. This objective is very
important as the material treated in this course
forms the foundation of plenty other subjects.
As a result, without a clear-cut and well-
presented course on signal processing, the stu-
dent may miss out on the understanding of
further courses in the curriculum. Hence, the
course must integrate the basic concepts of
linear signal processing in the time- and trans-
formation domain together with the biological-
ly inspired computational paradigms imple-
mented by artificial neural networks and the
signal processing algorithms running on kilo-
processor arrays. Besides delving into the theo-
retical foundations, the course tries to demon-
strate each principle by applications in the field
of information technologies (e.g. adaptive algo-
rithms for channel equalization and data com-
pression, pattern recognition and data mining
examples). In this way, the students can come
to grip with basic notions of signal processing
via applications as well.
Keywords: digital signal processing, neural
networks and computing
I.
I
NTRODUCTION
The two fundamental areas of signal pro-
cessing are (i) linear; and (ii) nonlinear signal
processing. Treating both of the areas in one
course has the advantage of addressing general
issues such as (i) representation; (ii) learning;
(iii) and generalization together. In this way,
the students are not only exposed to the de-
scription of signals in time- and transform
domains, but to the fundamental issues of sig-
nal processing as well. This helps to open up
new horizons of understanding not only in
linear but also neural based signal processing.
This can serve the orientation of students in
the info-bionics program who need to get an
insight into the signal processing paradigms
emerging neural based systems.
The third part of the course deals with kilo-
processor based signal processing which is an
important direction of present day signal pro-
cessing technologies. These technologies
make possible to integrate 5 billion transistors
in a single chip, which, however poses new
challenges to signal processing (taking into
account dissipation and signal propagation
delays). Thus it is important to highlight how
to implement the traditional algorithms on
these novel architectures.
The course materials (slides, dictionary,
etc.) have been designed to serve this integral
approach to signal processing.
II.
M
ATERIALS ALREADY AVAILABLE FOR
THE COURSE
Some of the topics touched upon in the
course can be found in the following books:
J.G. Proakis and D.G. Manolakis, Digi-
tal Signal Processing, Prentice Hall,
1996;
43