LDPC-based Joint Source Channel Coding and Decoding Strategies for single relay cooperative communications

Abstract

This paper investigates new cooperative communication strategies based on Low-Density Parity-Check codes (LDPC) with Joint Source-Channel (JSC) coding and decoding algorithms to efficiently deliver a correlated content to a destination node. The proposed strategies rely on double LDPC codes where the source and the channel coding are similar to standard LDPC operations, and can be mapped on Tanner graphs for message-passing decoding. Three coding strategies are proposed where different LDPC source and channel coding setups are applied at the source node and the relay node. For each transmission strategy, we propose a graph mapping the whole network, over which iterative message-passing decoding is applied, and where the probability of errors in the source–relay link are taken into account. We showed, based on computer simulations, that the proposed strategies can provide substantial improvements compared to equivalent rates point-to-point systems. Simulation results also showed that the system performance depends on the relay position, and that adapting the source and the relay channel coding rates accordingly can enhance the overall system performance for a time-varying network.