A Difference RSS-Based 5G Positioning Method With DALS Optimization

Difference of received signal strength (DRSS)-based positioning has commercial advantages due to its low system complexity. This article focuses on improving the positioning accuracy of DRSS by using multiple base stations (BSs). An improved two-step linear least squares (TLLS) estimator is proposed to reduce the computational time. A number of initial estimated locations are obtained by using different combinations of BSs. After analyzing the characteristics of the distribution of initial estimated locations, a dense area of location search (DALS) method is proposed to filter them. The error probability distributions of the coordinates of the filtered estimated locations are given. Based on the estimated coordinates and Pathloss exponent (PLE) of each filtered estimated location, the corresponding approximated Cramer-Rao lower bound (CRLB) is derived as the variance of the error random variable. A maximum likelihood estimator (MLE) is proposed to estimate the final location. A 5G-Advanced-based open dataset is used to validate the proposed method in this article. The simulation results show a significant improvement in the positioning performance of DRSS using the proposed optimization method, with the positioning error reduced by 95.7% for 90% of the samples and the average computation time reduced by 99.34%.