LDPC-based Joint Source-Channel-Network Coding for the Multiple Access Relay Channel

Abstract

In this work, we investigate the MARC (Multiple Access Relay Channel) setup, in which two Markov sources communicate to a single destination, aided by one relay, based on Joint Source Channel Network (JSCN) LDPC codes. In addition, the two source nodes compress the information sequences with an LDPC source code. The compressed symbols are directly transmitted to both a relay and a destination nodes in two transportation phases. Indeed, the relay performs the concatenation of the received compressed sequences to obtain a recovered sequence, which is encoded with an LDPC channel code, before being forwarded to the destination. At the receiver, we propose an iterative joint decoding algorithm that exploits the correlation between the two sources-relay data and takes into account the errors occurring in the sources-relay links to estimate the source data. We show based on simulation results that the JSCN coding and decoding scheme into a MARC setup achieves a good performance with a gain of about 5 dB compared to a conventional LDPC code.

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

“…In [29], authors considered double photograph-based JSCC LDPC codes for radiography image transmission, and provided channel SNR (Signal-to-Noise Ratio) improvements. In another context, a coding and decoding scheme based on Joint Source-Channel-Network (JSCN) LDPC coding was proposed in [31] for the Multiple Access Relay Channel (MARC), and demonstrated good BER performance improvements. The state-of the art of the applications enabled by protograph LDPC codes are studied in [32] including joint channel-and-physical-layer network coding.…”

Reliability Analysis of the Joint LDPC Decoding Algorithms over the Multiple Access Channels

“…Also, spatially coupled LDPC codes are utilized to multi-user detection over the MAC models to get advantages [19]. For the Multiple Access Relay Channels, a joint source-channelnetwork coding scheme is investigated in [20], which employs LDPC codes. A minimum error probability approach is studied in [21] to present a joint optimization method for two-user case in a non-orthogonal MAC.…”