A novel robust algorithm attenuating non-line-of-sight errors in indoor localization

In time of arrival (TOA) based indoor personnel positioning system, the human body mounted with positioning devices could cause non-line-of-sight (NLOS) propagation and further give rise to large ranging errors, thus reducing the accuracy of localization. A novel NLOS identification and mitigation method based on UWB signal power is proposed in this paper, which solves the problem that the performance of classical least squares (LS) algorithm severely degrades in NLOS environments. An improved self-learning LS localization algorithm is also introduced, overcoming the drawbacks of LS estimator that it requires at least three range estimates for an unambiguous solution. Furthermore, we demonstrate the proposed approach outperforms traditional LS algorithm with an increasing localization accuracy by 50% in NLOS scenarios.