1. IEEE Signal Processing Magazine
2. Signal Processing Digital Library*
3. Inside Signal Processing Newsletter
4. SPS Resource Center
5. Career advancement & recognition
6. Discounts on conferences and publications
7. Professional networking
8. Communities for students, young professionals, and women
9. Volunteer opportunities
10. Coming soon! PDH/CEU credits
Click here to learn more.
10 years of news and resources for members of the IEEE Signal Processing Society
When did you first come into touch with signal processing? What was your motivation of following a career in this domain?
In 1984 I obtained my first degree which was a five years diploma in Electrical Engineering. My favorite courses were quite diverse, from Electromagnetism and Antenna Theory to Automatic Control, System Theory and Stochastic Processes. At that time, Signal Processing (although it was already a recognized discipline) was not yet as clear and distinct part of the curriculum as it became a few years later. There were some relevant elective courses but most of the Signal Processing material was rather diffused in other courses, mainly in those related to communications and control systems. The first time I really came in touch with some signal processing aspects was during my diploma thesis (final year project). The thesis was concerned with digital implementation of a multichannel sensing system that was used for monitoring several critical parameters of the oil deposit in North Aegean sea. The work included a theoretical as well as a design and implementation part. The path from theory to practice was really amazing and I think it determined my attitude for the years to follow.In 1985, I pursued and gained (after written exams) a competitive PhD fellowship in Applied Mathematics. Having gained this fellowship, I had the right to seek for a subject and advisor to any University in Greece. Although Signal Processing was always in my mind, I was still undecided, until I met a young and aspiring assistant professor (Sergios Theodoridis).His “contagious” enthusiasm made me to decide almost instantaneously!
The first signal processing algorithm I ever implemented in software was a recursive least squares lattice algorithm applied to seismic signals. The reflection coefficients of the corresponding lattice filter are related to the earth layer parameters which are those actually needed when processing seismic signals (if the lumped parameter assumption holds true). The code was written in Fortran and run in a mainframe computer.Writing a code was a thrilling task then, because debugging was mainly done by inspection and intuition(!) and, also, the time you had to run the code in the mainframe computer was precious since it was shared between many users.
The first SP algorithm I implemented for a real-time application was a speech coding algorithm. It was based on the relatively new (at that time) concept of Linear Predictive Coding and the platform I used for implementation was a Digital Signal Processor with fixed point arithmetic. Programming in DSPs of the first generations was quite a challenge since the capabilities of the development environments were rather limited.
Apparently, the changes we saw are tremendous and of such a scale that only very gifted futurists could have imagined about three decades ago.Of course, in late 80s, signal processing was already a recognized and well-respected field.But it was still confined to relatively few areas such as, audio, speech, communications, geophysics etc. and it mainly dealt with measurable signals that represent a “tactile” physical quantity. Nowadays, as the well-known motto of SPS says, “Signal Processing is everywhere”. Indeed, we can find signal processing to any aspect of modern life, even to the most unexpected ones.
If I had to identify the most significant change I saw, then I would definitely say that it is the broadening of what we consider as “signal”. Now, it is not only a measurable or observable physical quantity but it could (almost) be any information-bearing entity that may appear in many different formats.That is, in many cases, we may “signalize” (allow me the neologism) several more abstract entities and then we can employ powerful signal processing theories and techniques to extract useful information and study underlying phenomena. We can think of several examples of this type in diverse areas as: network science, bioinformatics, linguistics, cognitive sciences, graphics, etc.
Similarly to the broadening of the concept of “signal” we are experiencing a broadening of the meaning of “processing”. The already rich “toolkit” of signal processing is further enriched with data driven methods and we are witnessing a very interesting and ongoing merging of signal processing and machine learning.
For the past few years my research interests are mostly focused on distributed signal processing and learning (SP&L). Although, the area has been an arena of intense research efforts by many groups worldwide, I strongly believe that there is plenty of uncharted ground.Recent developments in Internet-of-Things, Device-to-Device communications and Human-Machine-Interfaces give rise to new challenges for those working on distributed techniques for signal processing and machine learning.
Moreover, interesting research lines arise if we generalize the concept of distributed SP&L to include virtual machines in the Cloud environment.Achieving an optimal cloudification, from SP&L viewpoint, as well as coping with Cloud limitations (technical outages, security issues, susceptibility to attacks, delay constraints for “real-time” applications, etc.) are rather unexplored topics which may involve new interesting problems.
Another line of current research is related to signal processing for wireless communications, in particular in the mmWave band. Hardware constraints, related to the high frequency and bandwidth, necessitate the development of new signal processing techniques which will be based on proper partitioning of the involved operations to the analog and digital domains.
A stronger interaction between different societies is always mutually beneficial. We should keep in mind that societies correspond actually to sub-disciplines of the same wider discipline. They have been created because the body of knowledge increases steadily and the relevant community expands to unmanageable numbers (of people and activities). So, mainly for organizational reasons, and in order to increase efficiency, more and more societies are being created. If I counted correctly, IEEE has presently almost 40 different societies. But this does not mean, in my opinion, that Electrical and Electronic Engineering is actually split into 40 well-separated scientific sub-disciplines. So, we can easily identify clusters of societies with significant overlap, to a lesser or greater extent.
Especially in the cluster of Information and Communication Technologies, it is very hard to define borders and, in fact, there is no reason to define ones. There are new fields which evolve very fast thanks to the interaction between several societies and disciplines. As typical examples we could mention the emerged fields of Data Science and Big Data Analytics. It is clear that any attempt to narrow down their interdisciplinary character would considerably hinder their further development.
As signal processing experts, we very often have the chance to work side by side with people from completely different backgrounds. In such interdisciplinary collaborations we may find great research opportunities with significant impact to the society. In my opinion, the most interesting interdisciplinary projects are those in which we take part with an open mind and a mutual respect spirit, and not just as “craftsmen” who apply their tools to new applications. We need to go one step further and try to understand the problem under consideration. This process may reveal new scenarios we could never have imagined before, and for which new theories and tools could be developed.
It is very important for the new PhD student to think about the real motives that prompted him/her to pursue a PhD. There are several legitimate motives and corresponding PhD models. Indeed, there are PhD students with academic interests and purely scientific profile, others with more industry orientation and also those who pursue a PhD in order to increase their chances for a good job. If this is clear from the beginning of the PhD, then better decisions can be made in collaboration with the advisor.
Also, a new PhD student must realize that the whole process is time consuming, quite complex and not deterministic at all. There may be stressful and frustrating periods but also many moments of joy and fulfillness. But, with high probability, a well-motivated, well advised and hard-working PhD student will end up with a very good dissertation and will have gained plenty of new experiences. It is a common belief that the PhD years are among the most fruitful and rewarding years in life.
The signal processing community in Greece is very active with strong presence in the country and internationally, both in the academia and the industry sectors. The SPS Greece Chapter was founded more than two decades ago and ever since it has deployed several activities aiming at the local community. It maintains good relations with local chapters of other societies as well as with SPS chapters of neighboring countries. I have served as Chair of the Chapter for six years (i.e., for two terms) and I really enjoyed this period.I had the opportunity to collaborate with other colleagues and co-organize several events, including workshops, seminars, lectures by invited Distinguished Lecturers, etc.
The Chapter has set an ambitious plan for the next three years which comprises new activities that will further strengthen the bonds of the local community. A major goal is to increase the number of students who are engaged in chapter activities. More generally, motivating students and young engineers to become SPS members is indeed something that SPS should focus on more effectively in the near future.
Kostas Berberidis received the Diploma degree in electrical engineering from DUTH, Greece, in 1984, and the Ph.D. degree in signal processing and communications from the University of Patras, Greece, in 1990. During 1991, he worked at the Signal Processing Laboratory of the National Defense Research Center. From 1992 to 1994 and from 1996 to 1997, he was a researcher at the Computer Technology Institute (CTI), Patras, Greece. In period 1994/95 he was a Postdoctoral Fellow at CCETT/CNET, Rennes, France. Since December 1997, he has been with the Computer Engineering and Informatics Department (CEID), University of Patras, where he is currently a Professor, and Head of the Signal Processing and Communications Laboratory. Also, since 2008, he has been Director of the Signal Processing & Communications Research Unit of CTI.
He has been active in areas as: distributed processing and learning, adaptive signal processing, signal processing for communications, wireless communications and sensor networks, array signal processing, smart grid etc. He has published more than 160 papers in refereed international technical journals and proceedings of international conferences. He is co-author of four international patents as well as co-author of four technical books. He has been the principal investigator of more than 30 national, EU and bilateral projects.
Prof. Berberidis has served as a member of scientific and organizing committees of several high quality international conferences, as Associate Editor for the IEEE Transactions on Signal Processing, the IEEE Signal Processing Letters, the EURASIP Journal on Advances in Signal Processing and the ELSEVIER Pattern Recognition journal. Moreover, he has served as a Guest Editor for the EURASIP JASP and as member of the Best Paper Awarding Committee for EURASIP JASP Journal (from 2009 – today). He serves as a member of the Board of Directors of EURASIP, since January 2017. Also, from February 2010 until December 2017 he served as Chair of the Greece Chapter of the IEEE Signal Processing Society. He is a member of the “Signal Processing Theory and Methods” Technical Committee of the IEEE Signal Processing Society and the “Signal Processing for Communications and Electronics” Technical Committee of the IEEE Communications Society. Moreover, since August 2015 he is a member of the EURASIP Special Area Teams “Theoretical and Methodological Trends for Signal Processing” and “Signal Processing for Multisensor Systems”. He is a member of the Technical Chamber of Greece, a member of EURASIP, and a Senior Member of the IEEE.
© Copyright 2018 IEEE – All rights reserved. Use of this website signifies your agreement to the IEEE Terms and Conditions.
A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity.