Long-term Prediction Algorithm for Navigation Satellite Clocks Supported by Inter-satellite Links

Long-term prediction accuracy of navigation satellite clocks determines the autonomous navigation service performance. The long-term predictability of the clock deviation is limited by the long-term frequency stability of the satellite-borne atomic clock. The new-generation navigation satellites are planned or have been equipped with inter-satellite links to achieve autonomous navigation. With the support of the inter-satellite link, this paper uses the relative clock deviation measurement of the inter-satellite network to generate a virtual composite clock whose long-term frequency stability is better than all satellite-borne atomic clocks. The virtual clock deviation is predicted for a long time. Combined the predicted virtual clock and the real time estimation of clock deviation from the Kalman filter, the long-term prediction value of the clock deviation of each satellite is obtained, thereby improving the accuracy of the long-term prediction of the satellite clock deviation. The simulation results show that the method in this paper can make the prediction accuracy of the satellite clock deviation better than 20 ns after 30 days, which can be improved by more than 90% compared with the conventional single satellite clock deviation prediction accuracy.