Bio:

P. R. Kumar has worked on problems in game theory, adaptive control, stochastic systems, queueing networks, manufacturing systems, wafer fabrication plants, simulated annealing, machine learning, and information theory. His research is currently focused on power systems, wireless networks, and cyberphysical systems. He is a member of the National Academy of Engineering of the USA, and the Academy of Sciences of the Developing World. He was awarded an honorary doctorate by ETH, Zurich. He received the Outstanding Contribution Award of ACM SIGMOBILE, the IEEE Field Award for Control Systems, the Fred W. Ellersick Prize of the IEEE Communications Society, and the Donald P. Eckman Award of the American Automatic Control Council. He is a Fellow of IEEE. He is a D. J. Gandhi Distinguished Visiting Professor at IIT Bombay, and an Honorary Professor at IIT Hyderabad. He was a Leader of the Guest Chair Professor Group on Wireless Communication and Networking at Tsinghua University, Beijing, China. He was awarded the Daniel C. Drucker Eminent Faculty Award from the College of Engineering at the University of Illinois, the Distinguished Alumnus Award from IIT Madras, and the Alumni Achievement Award from Washington University in St. Louis.

Abstract:
The Poisson channel model is an apt representation of a direct-detection dark-current-limited optical communication link. The input to the channel controls the instantaneous rate of a Poisson point process at the photodetector output. Additionally, in a free-space optical communication link, atmospheric turbulence causes random variations in the refractive index of air at optical wavelengths, resulting in optical ``fading" that can be combatted by adaptive power control and MIMO techniques as in RF communication. This talk will survey the information theory of the Poisson channel, and implications for optical communication.

Bio:

Prakash Narayan received the B.Tech. degree in Electrical Engineering from the Indian Institute of Technology, Madras, and the M.S. and D.Sc. degrees in Systems Science and Mathematics, and Electrical Engineering, respectively, from Washington University, St. Louis, MO.

He is Professor of Electrical and Computer Engineering at the University of Maryland, College Park, with a joint appointment at the Institute for Systems Research. He has held visiting appointments at ETH, Zurich, the Technion, Haifa, the Renyi Institute of the Hungarian Academy of Sciences, Budapest, Universitat Bielefeld, LADSEB, Padova and the Indian Institute of Science, Bangalore.

Narayan's research and teaching interests are in multiuser information and communication theory, cryptography, broadband communication networks, and information theory and statistics. He has served as Associate Editor for Shannon Theory for the IEEE Transactions on Information. Theory and is currently on its Executive Editorial Board. He was a member of the Board of Governors of the IEEE Information Theory Society, and is a Fellow of the IEEE.

Go to top     Go to the technical program page →

Abstract:
Cellular communications as we know it today has been shaping the planet in an unprecedented way. Today the vast majority of people on the globe are connected due to cellular, and with current technology devices are becoming connected as well. Its success has flattened the planet and enabled economies to participate in the global economy, being the basis for many countries to become major players in world-wide trade and business.

So far we have seen that cellular enables content of different categories to be moved around the globe, where prominent examples of content are voice telephony, text messaging, video streaming, emails and files. The increase in volume and size of the content has driven the need for developing cellular standards that can handle continuously increasing data rates. However, is there another frontier to be tackled? Comparing the orders of magnitude of increase in data rate with the round-trip latency of interaction, the latter has not dropped much below the requirement for telephony. LTE achieves a typical round-trip latency of 25ms, still far beyond a maximum of 10ms to start enabling real-time wireless gaming. When moving to a round-trip latency of 1ms, and carrier grade robustness and availability, a new breakthrough in enabling unprecedented mobile applications becomes viable. These applications are coined "Tactile Internet", and will dramatically reshape our societies. The breakthrough is that we will be able to steer and control our real and virtual world around us with a tactile real-time interaction, without creating cyber-sickness. This will revolutionize education, mobility & traffic, health & care, sports, entertainment, gaming, and the smart grid, just to name some segments which can be seen today already.

To make this vision come true, the 1ms round-trip latency budget sets tight constraints on every element in the chain of communication and interaction. The physical and media-access control layer therefore can only contribute up to 100µs, which clearly shows that LTE with a basic TTI (transmission time interval) of 1ms and an OFDM symbol of approximately 70µs; is out of scope. Hence, a new physical layer needs to be created. On the one hand the benefits of multi-carrier should be preserved, but due to very short symbol durations we will see frequency selectivity per sub-carrier and inter-carrier interference. As the benefits of orthogonality are lost, alternative multi-carrier schemes must be found. We propose GFDM (Generalized Frequency Division Multiplexing), which allows for engineering signals packets in frequency and time, and for a transceiver implementations with little additional cost versus OFDM.

Abstract:
Information theory and signal processing have classically used the notion of "side information" to formally describe and analyze situations where providing more information to either the encoding or the decoding process improves system performance. We consider and extend this viewpoint to modern day imaging systems, where in addition to images representing visual information, devices also capture a variety of side information. In the ubiquitous smartphone, for example, multiple sensors (microphone, GPS, accelerometer, compass, etc.) augment the two cameras that have become the norm. Additionally, these devices are usually connected to a large network of digital data. This rich "side information" can improve the performance of imaging applications and enable completely new functionality. Using research examples from our group, ranging from near-infrared, location coordinates, big data to semantics, we present applications of how such "side information" enables functionality and improvements in computational photography and
computer vision applications.

Bio:

Prof. Sabine Süsstrunk (http://ivrg.epfl.ch/people/susstrunk) leads the Image and Visual Representation Group (IVRG) at the School of Information and Communication Sciences at Ecole Polytechnique Fédérale de Lausanne (EPFL) since 1999. Her research interests are in computational photography, multimedia, color image processing and computer vision, and image quality. She has authored and co-authored over 120 publications, of which 6 have received best paper/demo awards, and holds 7 patents. She received the IS&T/SPIE 2013 Electronic Imaging Scientist of the Year Award for her contributions to color imaging, computational photography, and image quality.