On Feasibility of Interference Alighment in MIMO Interference Networks and a New Training Protocol for Channel State Estimation in Wireless Relay Networks

Cenk M. Yetis
Istanbul Technical University

Monday, March 1, 2010 at 10:00am
Engr IV Maxwell Room 57-124

Abstract
In the first part of the talk, we explore the conditions required for interference alignment to be applied in a K-user MIMO interference network. Interference alignment is a recent popular scheme not only due to its ability to allocate 1/2 of the bandwidth for each user simultaneously (conventional schemes, e.g., TDMA and FDMA, can allocate 1/K of the bandwidth) but also due to its linearity (requires a simple design of transmit and receive beamforming vectors).

Cross-layer design with interference alignment by using the results of this work is an interesting future work. In addition, an extension of this work to X networks (every transmitter transmits to every receiver) and flexible X networks (every transmitter does not necessarily transmit to every receiver) is not trivial and more questions rise from cross-layer design aspect. In fact, bandwidth limit per user for flexible X networks is also an interesting future work from information theoretic aspect. Finally, an iterative algorithm for interference alignment in these networks is still an open problem. Approximate talk length is 35 minutes.

In the second part of the talk, we introduce a new training protocol for channel state estimation in wireless relay networks that provides a higher network capacity and a faster transfer of channel state information (CSI) than the existing training protocols.

Our approach enables us to obtain the effective SNR (eSNR) of a MIMO wireless relay network as a function of important parameters including the number of relays, antennas, and channel estimation errors. An interesting future work is the optimization of these and other important parameters for wireless relay networks including the training time, power, and structure in order to maximize the eSNR (thus the capacity).

Biography
Cenk M. Yetis received his B.Sc. degree in Electronics Engineering from Isik University, Istanbul, Turkey in 2001. He received his M.Sc. degree in Telecommunications Engineering from Istanbul Technical University (ITU), Istanbul, Turkey in 2004. He expects to receive his Ph.D. degree in Satellite Remote Sensing and Communication from ITU in 2010. He received The Turkish Scienti c and Technological Research Council of Turkey (TUBITAK) scholarship from 2005 to 2009. From 2003 to 2007, he was full-time employed by Avea, one of the top three wireless services providers in Turkey, where he held rotational responsibilities in operation and planning groups. Between 2007 and 2008, he was on leave at Ohio State University and since 2008, he is on leave at University of California Irvine as a visiting researcher. His research interests include signal processing, information theory, and communication theory for wireless communications.