An Efficient Quantized Message Passing Receiver Design for SCMA Systems

Sparse code multiple access (SCMA) has been considered as an efficient technique to provide both massive connectivity and high spectrum efficiency for future machine-type wireless networks. However, the conventional uniform quantization of the message passing algorithm (MPA) for the SCMA detection induces a significant bit error ratio (BER) performance degradation. In this sense, we propose a new quasi-uniform quantization scheme that can efficiently handle the dynamic range in the exchange of messages. To accelerate the convergence of conventional Max-log MPA, the Sub-log MPA is considered by using the latest updating messages at the current iteration. Simulation results show that the proposed quasi-uniform quantization method can significantly improve the BER performance of the SCMA decoder without modifying the resource nodes' and variable nodes' update rules in both the additive white Gaussian noise and the Rayleigh frequency selective channels, as compared to the uniform quantizer.