Local-Search Based Detector for Decode-and-Forward Protocol Cooperative Systems

In this paper, we propose a local search based detector for cooperative diversity systems. The considered system involves one source, one destination and multiple single-antenna relays. It adopts the decode and forward (DF) protocol where relays could commit errors in decoding the data. At the destination, we propose a detector based on the likelihood ascent search (LAS) approach. It is well known that LAS algorithm requires a maximum likelihood (ML) decoding metric that will be improved from one iteration to another to refine the solution. The main contribution of this work consists in providing a closed-form expression for the aforementioned ML decoding metric. We prove that the proposed detector dramatically reduces the computational complexity compared with the ML detector. However, numerical results show that, for PAM constellations, the proposed scheme has the same performance as the ML detector, and for QAM and PSK constellations, it has near ML performances.