Vision Channel Knowledge Map‐Aided mmWave Beam Selection in V2I Communications

ABSTRACT Vision‐aided millimetre wave beam selection has gained widespread attention because of its ability to reduce pilot overhead. However, its practical deployment faces challenges due to image transmission latency and high computational overhead. In this letter, we propose a novel environment‐aware communication method called the vision channel knowledge map (VCKM), which minimises redundant image information and improves beam selection performance. To evaluate the method, we use the ViWi colo‐cam scenario dataset, simulation results show that the VCKM‐aided beam selection has better performance than the existing solutions.