User Grouping and Power Allocation for Downlink NOMA System with Statistical CSI

In this paper, we consider a downlink non-orthogonal multiple access (NOMA) system over Nakagami-m fading channels. The base station (BS) serves several groups of users with hybrid multiple access scheme. A new user grouping and power allocation (UG-PA) strategy is proposed to enhance sum throughput. Based on the UG-PA strategy, we derive the exact and asymptotic closed-form expression of outage probability, and we further obtain the diversity order. Finally, we formulate a power allocation optimizaition problem to maximize sum throughput. We also design a genetic algorithm (GA) to find a suboptimal solution. Monte Carlo simulations verify the accuracy of our derived outage probability and diversity order. Comparisons with benchmark indicate the effectiveness of our proposed UG-PA strategy.