Joint optimization of user scheduling and power allocation based on mmWave-NOMA system

Abstract

Due to the interference caused by the superimposed transmission of data streams to users and the constraint on the number of reusable users in a beam, the problem of users' power allocation that affects the actual system performance in mmWaveNOMA system is NP-hard. The common approach in the literature for this problem is to solve power allocation and user scheduling respectively to obtain suboptimal results. In this paper, we proposed an efficient heuristic scheme named EHS which joint user scheduling and power allocation to maximize the weighted sum-rate (WSR) as the system performance objective. Firstly, we introduce a series of auxiliary functions to transform the non-convex objective equation into a convex function under certain conditions. Secondly, we divided the NP-hard problem into two resoluble sub-problems. The single beam user scheduling algorithm based on dynamic programming (SBUS) and the multi-beam power control (MBPC) algorithm based on polynomial knap-bag problem are proposed to solve the two sub-problems respectively. Finally, the SBUS algorithm and the MBPC algorithm are combined to optimize the ultimate goal. Numerical results show that the total rate of the system achieved by EHS is 13.6% higher than existing Fractional transmit power control algorithm (FTPC).