Position and Machine Learning-Aided Beam Prediction and Selection Technique in Millimeter-Wave Cellular System

In this paper, a position-based beam prediction and selection technique in mmWave cellular system is proposed. Here, the next beam is predicted and selected by leveraging the position (i.e., distance, angles-of-arrival and -departure) information of the high-velocity user. The achievable rates are formulated for the proposed approach and compared with the conventional sequential beam sweeping technique. Though the performance with the proposed position-based approach can be improved significantly, it may fail when a blockage occurs. The proposed position-based approach then leverages the machine learning tools to deal with the blockages.