Reliability-based detection for decode-and-forward relaying in cooperative communications

This paper considers the decode-and-forward (DF) relaying scheme in cooperative communication networks. Specifically, this study considers a cooperative network, in which each relay can obtain perfect channel state information (CSI) from the source to itself, and the destination can acquire perfect CSI from all the relays to itself, but the destination does not need all CSI from the source to all participating relays. For the considered semi-blind model, we propose a quantized reliability-relaying (QRR) scheme, in which each relay quantizes the value of the log-likelihood ratio (LLR) into three levels and forwards the message to the destination if the LLR value falls into the "send +1" and "send +1" regions, and keeps silent if the LLR value falls into the "no-send" region. Under the semi-blind model and the same average power constraint imposed at each relay, we compare the QRR scheme with the traditional DF scheme. Through the theoretical analysis, the work shows that the QRR scheme outperforms the traditional DF scheme. The simulation results also agree with the theoretical analysis.