GNSS Multipath Mitigation Using Channel Parameter Estimation Techniques
Date
2013-12-18
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Multipath propagation can pose significant challenges to satellite based navigation systems. It remains a dominant source of accuracy degradation and is a major issue for high precision GNSS applications. Multipath can result in biased GNSS measurements, which can lead to inaccurate position estimates or, through fading and self-interference, can cause loss of lock of the signals. Without accurate LOS delay estimation in multipath environments GNSS receivers cannot provide reliable positions, velocity and time (PVT) estimates. Although there are many algorithms proposed in the literature which endeavor to mitigate the effects of multipath, this research topic is still active as no final solution has yet been found.
Given the above, the problem of GNSS multipath mitigation is pursued in this work through the estimation of the parameters of multipath components. For this purpose, three different approaches are proposed and tested. First, a sequential ML-based approach is proposed that sequentially estimates the channel parameters with a smaller computational load compared to the conventional ML-based approaches. This approach uses a detection procedure to avoid over-estimating or underestimating the number of multipath components. For this reason, the proposed approach is more robust in dealing with severe multipath situations such as urban areas. Afterwards, this ML-Based approach is combined with a low-complexity Bayesian tracking algorithm to further decrease the computational load. In this way, the receiver switches between two modes of operation depending on the severity of the variations of the multipath channel. A set of simulation and data processing results is then used to assess the performance of this technique. The results show that the proposed system outperforms both the classical DLLs and the conventional ML-based algorithms. This algorithm is also used to characterize the distribution of the number of multipath components for some of the visible satellites in the collected data set.
Second, some of the most well-known adaptive filters (LMS, NLMS, RLS and APA) are modified and developed to be used for the purpose of equalization of the multipath channel. The very low computational load associated with these techniques make them more suitable for implementation in hand-held receivers. The innovative hard decision block used in the structure of their feedback procedure increases their efficiency. The presented simulation and data processing results show that the estimation performances of some of these techniques (RLS and APA) are comparable to near-optimal ML-based techniques at higher SNR values.
Third, the possibility of employing the Doppler shifted copies of the received signal in a fast fading channel for the purpose of improving the estimation performance of subspace-based methods is analyzed and tested through simulation and experimental results. The results demonstrate a considerable improvement in the estimation accuracy of the proposed system compared to the cases where diversity approaches are used.
Description
Keywords
Engineering--Electronics and Electrical