Multi-User Networks with Limited Feedback
|Event Date:||April 2, 2015|
|Speaker:||Dr. Erdem Koyuncu
|Speaker Affiliation:||University of California, Irvine
|Open To:||ACCEPTABLE FOR ECE 694
Limited feedback is an umbrella term describing the (preferably low-rate) transmission control information that is usually provided by the receiver(s) of communication systems. Not only limited feedback is a fundamental feature of many of today's wireless systems, it will also play a central role in tomorrow's networks as a key enabler for many emerging physical layer technologies such as
massive MIMO, interference management/alignment, and cooperative communications. In fact, most of these new technologies heavily rely on the availability of the channel state information (CSI) at the transmitters, which can only be provided by receivers' feedback. The design of practical limited feedback schemes, especially for multi-user networks, has thus become more important than ever.
One of the most distinctive features of multi-user wireless networks is the interference phenomenon, dealing with which becomes very important when shifting from the classical single-user paradigm of communications to a multi-user paradigm. In this talk, we present our work on the achievable performance gains in cooperative/non-cooperative multi-user networks suering from interference, and show how to design distributed limited feedback schemes to achieve these gains.
Conventional wisdom says the performance gains of a system with perfect transmitter CSI can be only achieved with innite feedback bits. This is due to the implicit assumption of using xed-length quantizers, where the receiver's feedback binary codewords all have the same length. In our talk, we also challenge the conventional wisdom by introducing variable-length quantizers in the feedback design and show that in fact it is possible to achieve the performance of a system with perfect transmitter CSI with only a few bits of feedback.
Erdem Koyuncu received the B.S. degree from the Department of Electrical and Electronics Engineering of Bilkent University, Ankara, Turkey in 2005. He received the M.S. and Ph.D. degrees in 2006 and 2010, respectively, both from the Department of Electrical Engineering and Computer Science of the University of California, Irvine, Irvine, CA, USA, where he is currently a postdoctoral
scholar. His research interests include communication and information theories, signal processing and quantization, limited feedback, interference channels, and more recently, network connectivity and stochastic geometry.