Ping-Pong Optimization of User Selection and Beam Allocation for Millimeter Wave Communications

Both millimeter wave (mmWave) communication and massive multiple-input multiple-output (MIMO) are important technologies in the 5G era. To reduce the cost of a mmWave massive MIMO system in practice, hybrid beamforming usually adopted, which however inevitability complicates both user selection and analog beam allocation. To this end, in this paper we jointly optimize user selection and beam allocation under a wideband frequency selective mmWave channel. To be practical, both beam collision and inter-user interference have been taken into account. To tackle the non-convexity of the formulated problem, we propose a ping-pong-like optimization method by using hybrid particle swarm optimization and simulated annealing (HPS). Concretely, the joint optimization problem is divided into two sub-problems and the near-optimal solution is approached via ping-pong iteration optimization. The Metropolis acceptance criterion of simulated annealing algorithm is introduced to overcome the drawback of traditional particle swarm optimization, improving global search capability of HPS algorithm. The simulation results verify the effectiveness and flexibility of the proposed method compared with existing methods.