Maximum likelihood based detector for PD-NOMA with statistical CSI: more efficient and lower complexity compared to SIC

Abstract

In this paper, we derive the maximum likelihood (ML) detector for downlink power domain non-orthogonal multiple access (PD-NOMA) in Rician fading channel, to enhance the detection performance of the previously proposed schemes. Then, we modify this ML detector to obtain the boundary based ML (BBML) detector which has much lower computational complexity compared to the original ML while it has the same error probability performance. This detector uses the full statistical channel state information (CSI), and for decision making, compares the received signal with the boundaries obtained based on ML criterion. The delay of this method is less than that of traditional successive interference cancellation (SIC). Analytic and simulation results show that the BBML detector is more efficient than SIC and also previously proposed multi-threshold detector (MTD), in downlink NOMA systems.