A NLOS Mitigation Algorithm for TOA Based Localization in Mixed LOS/NLOS Environments

In this paper, we discuss the time-of-arrival (TOA) localization problem in the mixed line-of-sight (LOS) and non-line-of-sight (NLOS) environment where the TOA measurements maybe corrupted by NLOS errors, which may severely degrade the localization accuracy. Under this condition, the LOS measurements are spatially consistent with source position, while the NLOS measurements corrupted by large and positive NLOS errors are always randomly divergent without consistency. Based on this principle, a soft-decision optimization method is developed to mitigate the NLOS errors. The proposed algorithm is heuristic, which can gradually highlight the LOS and suppress the NLOS in the optimization process along the consistent direction. It works well in mitigating NLOS errors, which can achieve the positioning accuracy close to that using only LOS measurements. The performance is analyzed from different aspects by simulations. The results show that the proposed algorithm performs better than the existing methods especially in the situation where NLOS error is large.