Mitigation of multipath effects in GPS and BDS positioning using window matching method based sidereal filtering

Abstract

Multipath effect is a main source of error in relative positioning, which cannot be eliminated or mitigated by differential algorithm. We discuss the topic of mitigating multipath at a static station in observation domain with GPS and BDS systems. At present, the sidereal filtering as one of the most commonly used multipath mitigating methods relies on the repetition period of multipath, which cannot be accurately estimated influenced by the maneuver of satellite orbit and the time interval of satellite ephemeris. The window matching method as a real-time method was proposed to reduce this effect. However, this method is affected by similarity measures in the process of real-time window matching. We propose a near real-time window matching method based on sidereal filtering. In the modified method, the satellite single difference residual is divided into segments and the cross-correlation method is used to obtain the multipath repeat time. At the same time, the second segment series overlaps with the previous segment to ensure a near real-time performance. Based on the obtained repeat time, the template window and matched window are formed by epochs in the segment and then an affine transformation is applied to determine the value of multipath correction between the two windows. Tests were conducted for GPS and BDS systems respectively using the baseline observations at static stations in the Ha-Jia high-speed railway. The experimental results show that the modified method can mitigate the multipath error in double difference observation, and finally provide higher positioning results than methods without model and traditional model. In practice, application of the modified method in near real-time baseline positioning can effectively mitigate the multipath error.