Performance Evaluation of Loose Beamforming Using Direct-Binary Search in Massive MIMO Systems

When considering beamforming in massive MIMO systems, the large number of antennas increases the amount of calculation at a base station. There is also a concern about the complexity of antenna circuits. Therefore, we have proposed loose beamforming in which the beam is formed by deciding whether or not to use the antenna for each user. However, an appropriate antenna selection method for the loose beamforming has not been established yet. In this study, we introduce a direct-binary search into the generation of the loose beamforming weights and evaluate the performance. Although the throughput is lower than one of other optimization methods such as a genetic algorithm, it can significantly reduce the computation time and can generate appropriate weights for a short time, thereby confirming the effectiveness of the proposed method.