摘要：

Shannon's classic source and channel coding theorems proved that random codes are optimal for point-to-point communication systems. However, in network communication systems, random codes are not always optimal. For example, Korner and Marton first showed in 1979 that low-complexity structured (e.g., linear) coding beats the best known random coding scheme for the binary two-help-one source coding problem. This has led to much recent focus on the performance gains of structured (e.g., lattice) codes over random codes in multiterminal settings. In this talk, we advocate a hybrid random-structured coding approach for network communications. Our hybrid approach brings together the best of two different worlds of random coding and structured coding. It is inspired by Ahlswede and Han's scheme for the general two-help-one source coding problem. We apply hybrid random-structured coding to the Gaussian two-help-one problem. Together with a new lower bound, we are able to provide for the first time partial sum-rate tightness result. We additionally show that hybrid random-structured coding performs within two b/s from the optimal sum-rate.  

 

简历：
Zixiang Xiong received his Ph.D. degree in electrical engineering from the University of Illinois at Urbana-Champaign in 1996. He is currently a professor in the Department of Electrical and Computer Engineering at Texas A&M University. His main research interest lies in signal and information processing over networks. Dr. Xiong received a National Science Foundation Career Award in 1999, an Army Research Office Young Investigator Award in 2000, and an Office of Naval Research Young Investigator Award in 2001. He also received the 2006 IEEE SIGNAL PROCESSINGMAGAZINE Best Paper Award. He served as an Associate Editor for five IEEE Transactions. He was the Publications Chair of ICASSP 2007, the Technical Program Committee Co-Chair of ITW 2007, the Tutorial Chair of ISIT 2010, and the Awards Chair of Globecom 2014. He is a fellow of the IEEE.