Two-dimensional Beam Selection for Multi-user Massive MIMO System Based on Multi-armed Bandit Algorithm.

A massive Multiple-input Multiple-output (MIMO) system for multiple users is important to achieve large capacity for future wireless communications. As the number of beams and users increases, appropriate beam selection becomes difficult because of a large number of searches. In this paper, we propose a simple 2-dimensional beam selection method using the information of the received power for multi-user massive MIMO systems and formulate the 2-dimensional beam selection method with the e-greedy method. The objective is to improve the system performance by independently selecting the beam that maximizes the received power of each user terminal. By the numerical analysis, the validity of the proposed method is confirmed and the relationship between the search rate and the achieved received power is clarified.