A satellite orbit maneuver detection and robust multipath mitigation method for GPS coordinate time series

Abstract

Multipath effect is the main factor limiting the high-precision positioning of Global Navigation Satellite System (GNSS) because it cannot be eliminated by double-difference observations or existing empirical models. Although the receiver antenna technology can reduce the multipath effect, it also brings the cost burden and cannot effectively solve the short-delay multipath error of carrier phase observation. Sidereal filtering (SF) is a common method used to mitigate multipath effect in static positioning mode. However, the coordinate time series may be contaminated by outliers, and satellite orbits may be affected by maneuvers. These will lead to inaccurate estimation of the Multipath Repetition Period (MRP), and even deteriorate the multipath mitigation effect of the SF. To solve these problems, a Sidereal Filtering method with the Satellite Maneuver Detection and Robustness (SFSMDR) is proposed. In this method, based on the Orbital Repeat Time Method (ORTM), the Multipath Repetition Period with the Satellite Maneuvers Detection (MRPSMD) is determined. Furthermore, considering the robustness of median filtering to outliers, the coordinate time series of the first day were preprocessed by the median filtering. After applying the proposed SFSMDR method, experimental results show that the proposed SFSMDR can effectively detect satellite maneuvers. In the absence of satellite maneuvers, there is no need to re-estimate MRP, which can save computational cost to a certain extent. When the coordinate time series on the first day is contaminated by outliers, the positioning accuracy of the MRPSMD for the east and north directions, is decreased by 2.26% and 95.73%, respectively. In contrast, the positioning accuracy of the SFSMDR is increased by 46.21% and 52.71%, respectively.