Site-specific real-time GPS multipath mitigation based on coordinate time series window matching

Multipath effect is one of the main errors in precise global navigation satellite system (GNSS) positioning. Although there are many GNSS multipath mitigation models available, each model has some limitations. We discuss the problem of mitigating low-frequency multipath in a specific site from the coordinate domain. At present, sidereal filtering is one of the most commonly used methods, in which the repetition period of multipath has to be estimated before multipath mitigation. However, in such a method, the maneuver of the satellite orbit and the long ephemeris update interval will cause inaccurate estimation of the repetition period. We propose a variation of the sidereal filtering method based on window matching. In that method, the window of coordinate time series formed by the current epoch and several previous epochs is used as the query window, and the matched window is searched in the time series of the previous day. The degree of similarity between windows is evaluated by similarity measures. Based on the coordinate point pairs in the matching window, multipath mitigation of the current epoch is performed by an affine transformation. The effects of different similarity measures in window matching are explored, and an early and late matching algorithm is proposed to solve the ambiguous problem of multiple matching sequences in integer-valued similarity metrics. Two sets of short baseline data were collected by different types of receivers in harsh environments to validate the method. The results show the feasibility and effectiveness of the proposed method for site-specific real-time multipath mitigation.