Matching Code-on-Graph with Network-on-Graph: Adaptive Network Coding for Wireless Relay Networks

We consider user cooperation in a wireless relay network that comprises a collection of transmitters sending data to a common receiver through inherently unreliable and constantly changing channels. Exploiting the recently developed technology of network coding, we propose a new framework, termed adaptive network coded cooperation (ANCC), as a practical, de-centralized, adaptive and efficient cooperative strategy for large-scale wireless networks. The key idea is to match network-on-graph, i.e. the instantaneous network topology described in graph, with the wellknown class of codes-on-graph, i.e. low-density parity-check (LDPC) codes and LDPC-like codes, to combat the lossy nature of wireless links and to adapt to the changing network topology. We demonstrate, through several ensembles of low-density generator matrix (LDGM) codes and lowertriangular LDPC codes, that the proposed framework can considerably increase the cooperation level and reduce the outage rate. Huge gains of some 20-40 dBs are achieved over the conventional repetition schemes! Index Terms User cooperation, relay networks, network coding, code-on-graph, low-density parity-check (LDPC) codes, low-density generator-matrix (LDGM) codes