Interference networks are information-theoretic models for communication systems in which several independent transmissions occur concurrently on a shared medium. The primary example is a wireless system, where spectrum scarcity makes it necessary for users to transmit at the same time in the same frequency band. Despite great practical relevance and significant research efforts, the quest for characterizing the optimal data rate tradeoff and corresponding transmission schemes for single-hop interference networks remains unsolved.
We make progress in this direction by taking a simple modular approach. Instead of optimizing the encoder and decoder functions jointly, we restrict encoding to random code ensembles with superposition coding and time sharing, and focus on optimizing the decoder side only. We characterize the optimal rate region and show that it is achieved by a unifying simultaneous nonunique decoding rule, regardless of the relative strengths of signal, interference, and noise.
As a corollary, we show that the Han-Kobayashi bound, the best known inner bound on the capacity region of the interference channel with two user pairs, cannot be improved merely by using the optimal maximum likelihood decoder instead of the typicality decoder that is conventionally employed.
This is joint work with Abbas El Gamal (Stanford) and Young-Han Kim (UCSD).
Bernd Bandemer is a postdoctoral scholar at the Information Theory and Applications (ITA) Center at the University of California, San Diego. He received his Ph.D. degree in Electrical Engineering from Stanford University in January 2012, and his Dipl.-Ing. (MS) degree in Electrical and Computer Engineering from Ilmenau University of Technology, Germany, in 2006. In 2003/04, he was awarded a German-American Fulbright fellowship which he spent at Purdue University. His research focuses on network information theory and wireless communications, in particular, on understanding interference and designing transmission schemes that optimally handle it.