Robust GNSS Signal Tracking in Multipath and Interference Environments

Abstract

This paper investigates the problem of navigation satellite signal tracking in the presence of multipath and interference. From the Bayesian filtering perspective, the signal tracking problem in multipath and interference environments is defined as estimating channel parameters for all incoming signals (i.e. line-of-sight (LOS) signal, multipath replicas, and interference signals) at the same time. To track all the parameters based on Bayesian filters, accurate information on the incoming signals (e.g. signal numbers and dynamics of timevariant parameters) is required. However, knowledge on the multipath and interference signals may not be accurately available in practice because they sometimes suddenly disappear or appear. The inaccurate knowledge on the multipath and interference signals leads to the mismatch between the true received signal and the signal model. Consequently, the accuracy of the LOS signal tracking can be degraded. To overcome these problems, a robust satellite signal tracking algorithm is presented based on a particle filter combined with a robust beamformer. With the help of beamforming, the unwanted signals can be suppressed in the received signal, and the received signal can be modeled in terms of only the parameters for the LOS signal. Hence, the signal tracking problem in multipath and interference environments can be solved by solely tracking the LOS signal, and the proposed algorithm has the robustness to the signal model mismatch from the inaccurate knowledge on the unwanted signals. A novel robust beamformer minimizing the mean squared error is also presented for the multipath and interference suppression. To evaluate the performance of the proposed algorithm, extensive computer simulations are conducted. The results are compared with the extended Kalman filter-based and particle filter-based tracking algorithms that track parameters of all incoming signals without beamforming.