Machine Learning Based Blockage Prediction in Mm-wave V2I Communications

This paper shows how to predict mm-Wave enabled V2I (Vehicle to Infrastructure) communication quality. These days, automated driving is widespread. However, there are blind spots that cannot be recognized by the sensors installed on the vehicles. To circumvent this, infrastructure i.e. RSU (Road Side Unit) which is equipped with camera, LiDAR, which is a sensor that can emit laser beams and uses the information from the reflected light to measure the distance to an object and the shape of the object on the vehicles and antenna can send information to the vehicles. This paper attempts to predict such V2I communication quality merely based on images from the RSU camera. We propose a comprehensive prediction method that focuses on V2I scenarios and Numerical analyses were conducted to predict changes in received power of mm-Wave enabled V2I communication links in a complete simulation environment.