Machine Learning-Based Beam Selection for V2X Communication

초록

In this paper, we develop a machine learning (ML) framework for vehicular networks that aid beam selection in the presence of the other road users. The frequency of beam selection is determined by the speed of the vehicle, the channel coherence time and blockages between the transmitter and receiver. The results shows that the selection overhead can be greatly reduced even in a high-speed communication scenario. Subsequently the system throughput can b by allocating more time to data transmission.