Optimized Signaling of Binary Correlated Sources over GMACs

TL;DR

This paper develops optimized binary signaling schemes for two correlated sources transmitting over Gaussian multiple access channels, exploiting source correlation to improve performance with joint decoding.

Contribution

It introduces a method to design binary pulse-amplitude-modulation constellations that leverage source correlation to enhance non-orthogonal multiple access channel performance.

Findings

Optimized constellations improve system performance.

Correlation exploitation reduces interference effects.

Numerical results validate the proposed design.

Abstract

This work focuses on the construction of optimized binary signaling schemes for two-sender uncoded transmission of correlated sources over non-orthogonal Gaussian multiple access channels. Specifically, signal constellations with binary pulse-amplitude-modulation are designed for two senders to optimize the overall system performance. Although the two senders transmit their own messages independently, it is observed that the correlation between message sources can be exploited to mitigate the interference present in the non-orthogonal multiple access channel. Based on a performance analysis under joint maximum-a-posteriori decoding, optimized constellations for various basic waveform correlations between the senders are derived. Numerical results further confirm the effectiveness of the proposed design.