Optimizing GNSS observation model with onboard receiver for LEO precise orbit determination

Precise orbits of Low Earth Orbit (LEO) satellites are the prerequisite for various precise applications with LEO satellite/constellations. Typically, centimeter-level LEO satellite orbit products can be obtained using onboard GNSS observation data. However, with the arrival of the 25th solar activity peak year and the cost control of commercial LEO satellites on manufacturing costs, traditional methods for onboard GNSS data processing need further improvement. In terms of data preprocessing, due to the severe ionospheric variations caused by solar activities, traditional cycle slip detection models are prone to frequent false detections of cycle slips during ionospheric active periods, leading to a decline in the accuracy of LEO satellite orbit determination. This study analyzes the variation characteristics of ionospheric disturbances, and proposes a polynomial fitting prediction method with ionospheric variation constraints, which can effectively distinguish cycle slips from ionospheric variations and improve the LEO satellite orbit determination accuracy under ionospheric disturbances. The results show that with the constraints of ionospheric variation, the RMS values of orbital errors for GRACE-C in along-track, cross-track, and radial components are improved by 11%, 17%, 6%, respectively. As for the optimization of GNSS observation model for low-cost LEO satellite, this study proposes a GNSS observation model considering the time-varying characteristics of onboard receiver biases, which can effectively enhance the stability of onboard receiver clock offset solution and ensure the accuracy of LEO satellite orbit as well. The results show that new model can reduce the discontinuities of arc-boundary for receiver clock from tens of nanoseconds to sub-nanosecond levels. In terms of frequency stability, the new model shows the similar short-term stability to the conventional model while notable improvements in medium- and long-term stability (beyond 10²s).