Information Combining for OFDM-based Relay Networks

In this paper, we present a method for calculating lower and upper bounds on the mutual information of a broadband relay network. The analysis is carried out for a simple network consisting of a source, one relay and the destination. However, it can be easily extended to larger networks. The relay operates in a half-duplex mode and performs decode-and-forward. In order to overcome the effect of multipath fading, orthogonal frequency division multiplexing (OFDM) is used on all physical links. The capacity of this network ca be analyzed by the max-flow min-cut theorem delivering an upper bound for the general case. For specific networks like the degraded relay channel, the achievable rate can be exactly determined. In order to investigate the tightness of these theoretical results, a capacity approaching repeat accumulate code is considered. By using the information combining technique that has been developed for the analysis of iterative decoding strategies, tight bounds on the achievable mutual information can be numerically computed. It is demonstrated that the upper bound of the max-flow min-cut theorem is nearly reached by this code and that the result for the degraded relay channel can serve only as a lose lower bound. In particular, the relay is also useful even in a scenario where it is not able to fully decode the received message.