NLOS 3-D Multi-Target Positioning Algorithm for Urban Corner Based on Multipath Information

This paper proposes a non-line-of-sight (NLOS) 3D multi-target positioning algorithm based on multipath information in urban corner scenarios. Firstly, the positions of all multipath ghost points are calculated using the multipath angle of arrival (AOA) and time of arrival (TOA). Secondly, an improved K-means algorithm based on the Gaussian kernel function is proposed to filter interference ghost points and determine the refined ghost point position. Finally, a multi-target position estimation algorithm based on multipath ghost point matching is proposed to complete the multi-target position estimation. Simulation results show that the proposed algorithm can achieve high-precision NLOS multi-target 3-D positioning even in the presence of certain measurement errors.