Distributed Incremental Cooperative Relaying with Quantized Feedback

Incremental relaying (IR) is known to improve the spectral efficiency of cooperative communication by using the destination feedback. In IR, the relay nodes are employed to assist the source transmission only if it is failed. When the relay nodes perform demodulate-and-forward (DmF), all of them are eligible to forward the source information. This is different from decode-and-forward (DF), where only the relay nodes with successful decoding will forward. In fact, employing all the relay nodes to forward not only increases the end-to-end delay but also incurs a low spectral efficiency due to information repetition. This paper considers DmF-based IR and proposes a distributed relay cooperation scheme that achieves full diversity using the quantized feedback from the destination node. The quantized feedback allows the relay node to make the cooperation decision locally and only a small amount of bandwidth is consumed for signaling. We theoretically analyze the average end-to-end bit error rate (BER) based on the cooperative maximum ratio combining (C-MRC). Numerical results demonstrate that the proposed scheme achieves full diversity and the BER improvement is significant when the the source-destination channel is weak.