A non-line-of-sight error mitigation algorithm in location estimation

TLDR

Location estimation of mobile phones is a key research area, hindered by measurement and non‑line‑of‑sight errors that arise from blocked direct paths and are considered a major challenge. The study develops an algorithm to mitigate unidentifiable NLOS errors in mobile phone location estimation. The residual weighting algorithm (Rwgh) uses an over‑determined set of range measurements, and its performance is compared to two generic algorithms across deterministic and random NLOS error scenarios. Rwgh effectively mitigates NLOS errors in location estimation across tested scenarios.

Abstract

The location estimation of mobile telephones is of great current interest. The two sources of range measurement errors in geolocation techniques are measuring error and non-line-of-sight (NLOS) error. The NLOS errors, derived from the blocking of direct paths, have been considered as a killer issue in the location estimation. In this paper we develop an algorithm to mitigate the NLOS errors in location estimation when the range measurements corrupted by NLOS errors are not identifiable. This algorithm, referred to as the residual weighting algorithm (Rwgh), assumes that the number of range measurements is greater than the minimum number required. The Rwgh algorithm is compared with two other generic algorithms and tested under various scenarios and NLOS error models, which includes deterministic and random NLOS error models. In both cases, Rwgh has shown the effectiveness of NLOS error mitigation in location estimation.