Integrated precise orbit determination of GPS, BDS, and partial satellites selected from LEO constellation

Abstract

Traditional precise orbit determination (POD) for Global Navigation Satellite System (GNSS) typically relies on a considerable number of global ground stations. The rapid development of large Low Earth Orbit (LEO) communication constellations in recent years presents an opportunity for integrated POD of LEO and GNSS satellites. The participation of abundant observations from onboard GNSS receivers in GNSS POD can significantly enhance the orbit precision, availability, and integrity of GNSS satellites. However, the improvement in orbit precision comes with a notable increase in computation time when a substantial number of LEO satellites are involved. In this study, we propose a method for selecting LEO satellites based on the LEO geometric dilution of precision (LGDOP) value. By identifying the LEO satellite combination corresponding to the minimum LGDOP, partial LEO satellites evenly distributed are selected from the entire LEO constellation. Using the semi-realistic simulation approach, only onboard observations of the LEO constellation are simulated to enhance GPS and BDS POD. The average accuracy of orbit determination for simulated LEO satellites is 2.27 cm in 3D directions, comparable to FengYun-3 but inferior to GRACE-FO. When onboard observations from 10 LEO satellites, selected using the minimum LGDOP method, participate in the integrated POD, the orbit precision can be improved by 10 %-14 % compared with the sequence selection method. Consequently, the LGDOP selection method optimizes the distribution of LEO satellites and demonstrates superior effectiveness in enhancing GPS and BDS POD with an equivalent number of LEO satellites. Involving 20–30 LEO satellites in the solution can enhance the orbit precision of GPS and BDS satellites by approximately 40 % and 80 % for global and regional ground stations, respectively. Increasing the number of participating LEO satellites further does not significantly improve the GPS and BDS orbit precision but does notably increase the computation time, which exceed 12 h for regional stations and 14 h for global stations. Thus, selecting approximately 20–30 LEO satellites using the proposed method effectively balances orbit determination precision and computation efficiency.