|Abstract:||An important application of the Global Navigation Satellite System (GNSS) in space is the precise orbit determination (POD) of satellites, particularly that of Low-Earth Orbit (LEO) satellites. There are hundreds of LEO satellites currently in orbit; and, in the near future, the number of the LEO satellites will very likely reach into the thousands. How to efficiently and precisely determine the orbits of a larger number of satellites to satisfy mission accuracy and time requirements will become a challenging problem. Currently, GNSS-based POD of LEO satellites can be performed using either undifferenced (UD) or double-differenced (DD) observations. The main distinction between DD and UD POD involves the use of clocks and reference receivers. The UD POD requires both precise GNSS satellite orbits and clocks; while the DD POD requires not only the precise orbits, but also reference stations. Therefore, GNSS-based LEO POD is conventionally based on either global ground stations or precise GNSS clocks – both of which being inconvenient – for near real-time or real-time data processing. Because of this inconvenience, we investigated the feasibility of achieving adequate GNSSbased POD of LEO satellites without use of reference stations and precise GNSS clocks. Precise GNSS satellite orbits can be easily predicted in a certain time period. Conversely, precise clocks are very difficult to predict. The use of DD observations can avoid such a clock problem, which is particularly important for near real-time and real-time data processing. However, if the DD observations are performed by two GNSS satellites and two LEO satellites without global ground stations, only the precise relative (not absolute) orbits can be determined. What would happen if the DD observations are based on more than two LEO satellites, or perhaps a network of LEO satellites? To answer this question, this paper describes our findings regarding how well the LEO satellite orbits can be determined using DD measurements from a network of 36 LEO satellites that fly in the same inclination and plane with nearly the same separations. Also described is our proposed approach, including verification of it using simulated data. The results show that orbit accuracy at the centimeter level can be achieved, and is improved with the increment of LEO satellite number. Based on the proposed approach for near real-time and real time data processing, we conclude that the orbits -- both absolute and relative precise – for LEO satellites can be determined without use of global reference stations and precise GNSS clocks.|
Proceedings of the 30th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS+ 2017)
September 25 - 29, 2017
Oregon Convention Center
|Pages:||1747 - 1753|
|Cite this article:||
Kang, Zhigui, Tapley, Byron, Bettadpur, Sinivas, "GNSS-Based Precise Orbit Determination of LEO Satellites Using Double-Differenced Observations," Proceedings of the 30th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS+ 2017), Portland, Oregon, September 2017, pp. 1747-1753.
ION Members/Non-Members: 1 Download Credit