Transmitter beamforming control based on firefly algorithm for massive MIMO systems with per-antenna power constraint

Abstract

In the massive multiple-input multiple-output (MIMO) downlink, transmitter beamforming (BF) to achieve a beamforming (power) gain and/or a reduction in inter-user interference in a multiuser MIMO case based on the channel state information (CSI) at the transmitter side is essential. Massive MIMO requires a large number of transmitter power amplifiers dedicated to each of the transmitter antennas, and the allowable transmit power per power amplifier (thus per transmitter antenna) is severely limited. Therefore, it is necessary to perform transmission BF considering the limitation on the transmit power per antenna. In this paper, we consider BF control using the firefly algorithm (FA) assuming the per-antenna transmit power constraint aiming at maximizing the system throughput defined as log-sum user throughput (equivalently the geometric mean user throughput) corresponding to proportional fairness (PF). The FA is a kind of swarm intelligence and by mimicking the behavior of fireflies that move toward neighboring fireflies based on the distance and brightness of the neighboring fireflies, a multimodal optimization solution is obtained efficiently. By setting the BF matrix as the position of the firefly and the obtainable metric for the system throughput maximization as the brightness of the firefly, we apply the FA to the BF control. We evaluate the performance of BF control using the FA from the viewpoint of the geometric mean user throughput based on computer simulations.