Distributed Radio Map Reconstruction for 5G Automotive

Abstract

Radio maps are expected to be an essential tool for the resource optimization and management of 5G automotive. In this paper, we consider the problem of radio map reconstruction using a wireless sensor network formed by sensor nodes in vehicles, access nodes from a smart city infrastructure, etc. Due to limited resource constraints in sensor networks, it is crucial to select a small number of sensor measurements for field reconstruction. In this context, we present a novel distributed incremental clustering algorithm based on the regression Kriging method for radio map reconstruction in terms of average received power at locations where no sensor measurements are available. The path-loss and shadowing components of the wireless channel are separately estimated. For shadowing estimation, clusters of sensors are adaptively formed and their size is optimized in terms of the least number of sensors by minimizing the ordinary Kriging variance. The complexity of the proposed algorithm is analyzed and simulation results are presented to showcase the algorithm efficacy to field reconstruction.