Computer Vision Aided mmWave Beam Alignment in V2X Communications

• DOI: 10.48550/arxiv.2207.11409
• Publication Date: 2022-01-01
• Authors: Weihua Xu; Feifei Gao; Xiaoming Tao; Jianhua Zhang; Ahmed Alkhateeb
• Keywords: Transceiver

Visual information, captured for example by cameras, can effectively reflect the sizes and locations of environmental scattering objects, thereby be used to infer communications parameters like propagation directions, receiver powers, as well blockage status. In this paper, we propose a novel beam alignment framework that leverages images taken cameras installed at mobile user. Specifically, utilize 3D object detection techniques extract size location information dynamic vehicles around user, design deep neural network (DNN) optimal pair transceivers without any pilot signal overhead. Moreover, avoid performing too frequently or slowly, coherence time (BCT) prediction method is developed based on vision information. This improve transmission rate compared with approach fixed BCT. Simulation results show proposed methods outperform existing LIDAR solutions, demand much lower hardware cost communication