Mobile Positioning System Based on Virtual Base Station Transform and Convex Optimization

Abstract

Mobile positioning in the cellular systems usually suffers great performance degradation caused by the non-light-of-sight (NLOS) effect in the urban area. Thus, this work proposes a novel virtual base-station transform (VBST) algorithm based on the particle filter and convex optimization. The proposed algorithm can virtually move the base stations to the light-of-sight (LOS) positions to the mobile station without the limitation of the physical obstacles such as buildings and hills. To track the position of the mobile station, this study also presents an adaptive control mechanism to determine the positions of the virtual base-stations. The simulation results show that the positioning accuracy can be improved by about 30-meter root-mean-square-error (RMSE) under the fast NLOS fading environment.