Distributed Hypothesis Testing Over Multi-Access Channels

A coding and testing scheme is proposed for distributed hypothesis testing over multiple-access channels (MAC). The scheme combines hybrid coding with a MAC-version of Borade's unequal error protection. The resulting error exponent recovers the previous exponents by Rahman-Wagner and by Sreekuma-Gündüz for the special case of testing against conditional independence, and it is optimal when the sources at transmitters are independent under both hypotheses. In all these special cases, without loss in performance, hybrid coding can be replaced by separate source and channel coding.

[1]  Sergio VerdÂ,et al.  Statistical Inference Under Multiterminal Data Compression , 2000 .

[2]  Michèle A. Wigger,et al.  Distributed Hypothesis Testing Over a Noisy Channel , 2018 .

[3]  Shashibhushan Prataprao Borade,et al.  When all information is not created equal , 2008 .

[4]  Thomas M. Cover,et al.  Elements of Information Theory , 2005 .

[5]  Sung Hoon Lim,et al.  A Unified Approach to Hybrid Coding , 2015, IEEE Transactions on Information Theory.

[6]  Abbas El Gamal,et al.  Network Information Theory , 2021, 2021 IEEE 3rd International Conference on Advanced Trends in Information Theory (ATIT).

[7]  A. Wagner,et al.  On the Optimality of Binning for Distributed Hypothesis Testing , 2010, IEEE Transactions on Information Theory.

[8]  S. Amari,et al.  Error bound of hypothesis testing with data compression , 1994, Proceedings of 1994 IEEE International Symposium on Information Theory.