The effect of network coding on error performance in a noisy cooperative relay network

This paper investigates the effect of network coding on the error performance of a two user, single destination cooperative relay network in a flat Rayleigh fading environment. Each user acts as the relay node of the other user in forwarding the information to the joint destination. The users employ demodulate-and-forward and GF(2) and GF(4) network coding techniques. The destination demodulates the incoming data and decodes the network code using a Tanner graph. The maximum likelihood (ML) rule for decoding the network code is derived and used in analyzing the error performance of the system. The simulation results show that it is possible to achieve a diversity order close to 2 with GF(2) network coding and a diversity order close to 3 with GF(4) network coding. Hence, using GF(4) instead of GF(2) network coding significantly improves the error performance of the relay network under consideration, by increasing the diversity order.