Correlation between Channel State and Information Source with Empirical Coordination Constraint

TL;DR

This paper explores how the correlation between channel state and source symbols affects the joint source-channel coding, focusing on empirical coordination and lossless transmission, and demonstrates that such correlation enhances transmission feasibility.

Contribution

It characterizes achievable joint distributions under empirical coordination constraints and shows how source-channel correlation improves transmission feasibility.

Findings

Correlation improves the feasibility of lossless transmission.

Achievable joint distributions are characterized under empirical coordination.

Correlation between source and channel state reduces minimal distortion.

Abstract

Correlation between channel state and source symbol is under investigation for a joint source-channel coding problem. We investigate simultaneously the lossless transmission of information and the empirical coordination of channel inputs with the symbols of source and states. Empirical coordination is achievable if the sequences of source symbols, channel states, channel inputs and channel outputs are jointly typical for a target joint probability distribution. We characterize the joint distributions that are achievable under lossless decoding constraint. The performance of the coordination is evaluated by an objective function. For example, we determine the minimal distortion between symbols of source and channel inputs for lossless decoding. We show that the correlation source/channel state improves the feasibility of the transmission.