Particle Filter-based GNSS Positioning with NLOS Multipath Detection

Abstract

This paper proposes a novel positioning technique that can be used in vehicles and by pedestrians in urban environments in which large global navigation satellite system (GNSS) positioning errors occur because of multipath signals. In GNSS positioning, invisible satellites that are obstructed by buildings emit reflection and diffraction signals, which are called non-line-of-sight (NLOS) multipath signals. These cause major positioning errors in GNSS positioning. Thus, to mitigate NLOS multipath errors, NLOS signals must be identified from all the received GNSS signals. The proposed novel positioning technique can detect NLOS signals without using any external data and sensors to improve positioning accuracy in urban environments, where NLOS signals cause major positioning errors. The key idea behind this work is estimating the user’s position using the likelihood of position hypotheses computed from GNSS pseudoranges, which consists of only LOS signals, based on the analysis of pseudorange residuals. To determine the NLOS GNSS signals from the pseudorange residuals at the user’s position, it is in turn necessary to accurately determine the position before the computation of the pseudorange residuals; we solve this problem by using a particle filter. We propose a likelihood estimation method using the Mahalanobis distance between the hypotheses of the user’s position and these particles. The likelihoods can be evaluated using only the LOS pseudoranges determined from the pseudorange residuals. To confirm the effectiveness of the proposed technique, a positioning test was performed in a real-world urban-canyon environment. Using the proposed method, the distribution of the particles converged to within 1 m of the true position after few iterations of resampling. The proposed method is suitable and effective for accurately estimating the user’s position in urban canyons, where conventional GNSS positioning can cause large positioning errors.