Compressive Soft Forwarding in Network-Coded Multiple-Access Relay Channels

In this paper, we design a novel estimate-and-forward (EF) protocol in multiple-access relay channels (MARCs) where two correlated sources transmit their information to the common destination with the help of a relay. With the application of network coding (NC) at the relay, the network-coded soft symbols contain redundant information due to the correlated sources, which can be compressed before being forwarded. Specifically, we first derive the value of the network-coded soft symbol by considering the correlation of the two sources. Due to this correlation, the majority of soft symbol values are close to one. Then, we convert network-coded soft symbol vector into a sparse vector, which is suitable for compression by using the compressive sensing (CS). Next, we analyze the performance of the proposed EF protocol by comparing the soft symbol value and received signal-to-noise ratio (SNR) with the conventional EF protocol. Simulations show that our protocol can achieve the same bit error rate (BER) as the conventional EF protocol with proper compression rates, and it consumes much less transmission time due to the compression.