GNSS multipath estimation and mitigation based on particle filter

Abstract

Multipath interference is the main error source for high-precision positioning applications, and multipath suppression and mitigation is particularly important. Particle filters are widely used to solve non-linear filtering problems without limitation of Gaussian distribution, and Global navigation satellite system (GNSS) multipath estimation and mitigation based on particle filter are proposed in this study. This approach has four innovations: Firstly, the Kalman-based multipath signal model is improved to obtain a particle filter multipath signal model, and I and Q two channel signals in the algorithm model solve the coherence correlation phase estimation problem. Secondly, the particle filter algorithm constructs the correlation function using the current particle amplitude and delay. Thirdly, taking full advantage of every particle information, the weighted average is used to calculate state quantity of multipath. Fourthly, three resampling algorithms, including simple random, pseudo-parallel genetic algorithm and niche genetic algorithm, are taken to resample. Particle filter algorithm for multipath estimation and mitigation based on simulated data and actual navigation satellite signal data is verified. The simulation results show the proposed algorithm has high accuracy in multipath estimation than MEDLL and TK-MEDLL, especially in multichannel multipath estimation. The actual experimental results show particle filter algorithm is effective for improving the positioning accuracy in complex environments.