Indoor Sparse NLOS Environment Localization Method Based on Compressed Sensing

High-precision location-based services (LBS) have become a crucial infrastructure for modern society, yet identifying and suppressing non-line-of-sight (NLOS) error is a significant factor that limits positioning accuracy. To tackle the challenge of achieving high-precision localization without prior information, this paper proposes an indoor NLOS environment localization method based on compressed sensing. Firstly, the location of the user equipment (UE) is estimated using an improved 2.5D Chan algorithm, and subsequently, a discrete transformation method with time-difference-of-arrival (TDOA) observations to sparse NLOS space is constructed based on the estimated locations. And using the reconstructed sparse NLOS error vector, the localization error can be further reduced by an iterative method with adaptive thresholding (IMAT). The simulation results demonstrate that the proposed method displays higher accuracy and robustness compared to traditional NLOS mitigation algorithms.