Minimum Error Probability Cooperative Relay Design

In wireless networks, user cooperation has been proposed to mitigate the effect of multipath fading channels. Recognizing the connection between cooperative relay with finite alphabet sources and the distributed detection problem, we design relay signaling via channel aware distributed detection theory. Focusing on a wireless relay network composed of a single source-destination pair with L relay nodes, we derive the necessary conditions for optimal relay signaling that minimizes the error probability at the destination node. The derived conditions are person-by-person optimal: each local relay rule is optimized by assuming fixed relay rules at all other relay nodes and fixed decoding rule at the destination node. An iterative algorithm is proposed for finding a set of relay signaling approaches that are simultaneously person-by-person optimal. Numerical examples indicate that the proposed scheme provides performance improvement over the two existing cooperative relay strategies, namely amplify-forward and decode-forward