Eur. Phys. J. B 64, 625-631 (2008)
https://doi.org/10.1140/epjb/e2008-00096-x

Mean field theory of communication

Statistical-physics approach to wireless communications

Graduate School of Informatics, Kyoto University, Sakyo-ku, Kyoto-shi, Kyoto, Japan

Corresponding authors: ^a tt@i.kyoto-u.ac.jp - ^b knakamur@i.kyoto-u.ac.jp

Received: 28 August 2007
Revised: 2 February 2008
Published online: 7 March 2008

Abstract

We discuss a Gaussian vector channel with random channel matrix, as a mathematical model for various wireless communication systems including multiple-input multiple-output (MIMO) and code-division multiple-access (CDMA) channels, and how statistical-physics approaches can be applied to analyze the Gaussian vector channel. We then present a mean-field description of the Gaussian vector channel in terms of a joint distribution of any single element of originally sent channel input and its estimate via posterior probability, which is a refinement of the “decoupling” result reported by Guo and Verdú for a set of statistically equivalent elements of channel input, the number of which is extensive.