Evidential Obstacle Learning in Millimeter wave D2D communication using Spatial correlation

The 5G technology uses device to device (D2D) communication in the millimeter waves (mmWaves) to support the increasing demand for higher data rates by the user equipments (UEs). Though millimetre wave can immensely increase the data rates, it suffers from extremely high propagation as well as penetration losses from obstacles, requiring almost a line of sight (LOS) path between a pair of communicating devices. To reduce such losses, communication links can be broken into multiple hops by using intermediate devices as relays, avoiding obstacles and thereby extending the range. Knowing the location and size of obstacles is key to choosing good relays. Satellite imagery can be used to do the same. But small obstacles like lamp posts, banners, poles and trees cannot be captured properly using satellite imagery. Moreover, satellite information is proprietary and costly. Our work concentrates on mapping obstacles in a given area to help in the process of relay selection. We propose a learning based strategy, which considers spatial correlation of obstacles, to build the obstacle map efficiently without relying on satellite imagery. We propose an evidential framework to model the confidence in the knowledge gained for each cell, which can model uncertainty better than a typical probabilistic model. We also propose an operation using Gaussian smoothing, to utilize the information available from the nearby cells. We have proposed a visibility graph based relay selection algorithm which reduces the overhead due to repeated updation of the location information by the UEs, and thereby improves the overall coverage dramatically. Through simulations, we show that our evidential framework based approach can learn the map faster and more accurately than the typical probabilistic model.