Evolutionary multi-objective optimization for RIS-aided MU-MISO communication systems

We study a multi-user multiple-input single-output downlink system aided by a reconfigurable intelligent surface (RIS). Users’ sum rate and transmit power are two important performance indicators in such systems. However, most existing works only optimize one of them, resulting in severe performance degradation of the other. Motivated by this, in this paper, we formulate a multi-objective optimization problem to maximize the sum rate of users and minimize the transmit power simultaneously. According to our early work on fitness landscape analysis of sum rate maximization problems, the proposed problem is inferred to be multi-modal. To solve this non-convex and multi-modal problem, we propose a novel multi-objective evolutionary hybrid beamforming (MEHB) framework to find different trade-off solutions between the two conflicting objectives. In particular, we employ different kinds of multi-objective evolutionary algorithms and multi-modal multi-objective evolutionary algorithms as the baseline of MEHB framework, so as to design the passive beamforming. And the active beamforming at the base station is optimized by the classical zero-forcing method. The simulation results have verified the effectiveness of the dominance-based evolutionary algorithms in handling hybrid beamforming problems.