University of California, Riverside

Department of Electrical and Computer Engineering

Asymptotic Universal Optimality in Cooperative Wireless Networks and in Delay-limited...

Asymptotic Universal Optimality in Cooperative Wireless Networks and in Delay-limited...
Petros Elia
Department of Electrical Engineering, University of Southern California

Date: Monday, February 27, 2006
Time: 1:00 pm
Location: Bourns Hall A265

The work relates to wireless cooperative networks as well as delay-limited communications with or without feedback.

In regards to cooperative wireless networks, outage-based asymptotic optimality results were known only for infinite duration networks, full knowledge of the channel and infinite decoding and signaling complexities. We have achieved the same optimality, for finite and minimum delays, small decoding complexity and without requiring knowledge of the channel at the intermediate relays. This was achieved for the most general network topology and statistical channel characterization.  Utilizing cyclic-division and commutative algebras of normal matrices, the same optimal performance is maintained with minimal signaling complexity, and even under maximal symbol asynchronicity.  An information theoretic consequence of the existence of the above schemes is the fact that the minimum-delay non-dynamic selection-decode-and-forward, the non-dynamic amplify(receive)-and-forward strategy with knowledge of the average communication rate at the relays, the dynamic amplify(receive)-and-forward and the minimum-delay dynamic-decode-and-forward cooperative diversity schemes have asymptotically equal outage regions.  In practice, our schemes allow for all the strategies to jointly exhibit near optimal performance with small network structure.

In regards to delay-limited communications with or without feedback, the work seeks to optimally utilize the benefits of multiple input, multiple output channels which promise substantial improvement in the rate and error performance for wireless mobile telephony. In a recent seminal paper though, Zheng and Tse accentuated the strong sub-optimality of the existing delay limited MIMO schemes but then proceeded to suggest that there exist some random Gaussian codes that achieve the fundamental rate-error performance tradeoff by asymptotically meeting the outage region of the channel. Their result sparked considerable worldwide interest in finding such codes.  We present the only explicitly constructed unified family of communication schemes that meet the theoretical performance limits placed by the high-SNR outage for any point-to-point coherent channel, independent of fading statistics, temporal noise correlation and number of antennas. These constructions are based on special cyclic division algebras, and their existence in fact allows for improvements on the known theoretical optimality limits. Practical substantiation was provided by our constructing the unified family of `perfect space-time schemes' which exhibit near optimal performance for all ranges of spectral efficiency and SNR. Their optimality is maintained in the presence of feedback and it carries over to the MIMO-ARQ as well as the antenna-selection channel, providing improvements in non-ergodic performance and hardware complexity. 

About the speaker:

Petros Elia received his B.Sc. in electrical engineering from the Illinois Institute of Technology, his M.Sc. in electrical engineering from the University of Southern California, and in May 2006 he will receive his Ph.D. degree in electrical engineering, at the University of Southern California under the guidance of P. Vijay Kumar. Between his B.Sc. and M.Sc. degrees he was a software consultant for companies like Merck and Procter-&-Gamble. Currently his main research interests are in the area of wireless networks.
More in Colloquia

More Information 

General Campus Information

University of California, Riverside
900 University Ave.
Riverside, CA 92521
Tel: (951) 827-1012

Department Information

Electrical and Computer Engineering
Suite 343 Winston Chung Hall
University of California, Riverside
Riverside, CA 92521-0429

Tel: (951) 827-2484
Fax: (951) 827-2425
E-mail: E-mail/Questions