Exponentially Weighted Particle Filter for Simultaneous Localization and Mapping Based on Magnetic Field Measurements

This paper presents a simultaneous localization and mapping (SLAM) method that utilizes the measurement of ambient magnetic fields present in all indoor environments. In this paper, an improved exponentially weighted particle filter was proposed to estimate the pose distribution of the object and a Kriging interpolation method was introduced to update the map of the magnetic fields. The performance and effectiveness of the proposed algorithms were evaluated by simulations on MATLAB based on a map with magnetic fields measured manually in an indoor environment and also by tests on the mobile devices in the same area. From the tests, two interesting phenomena have been discovered; one is the shift of location estimation after sharp turning and the other is the accumulated errors. While the latter has been confirmed and investigated by a few researchers, the reason for the first one still remains unknown. The tests also confirm that the interpolated map by using the proposed method improves the localization accuracy.