PRECISE TIME SYNCHRONIZATION OF 2 MILSTAR COMMUNICATIONS SATELLITES WITHOUT GROUND INTERVENTION

Satellite navigation and communication systems often require precise s ynchronization among spacecraft clocks. In the traditional method for achieving synchronization, a ground station makes time-offset measurem ents to the various spacecraft clocks, and then updates the time and f requency of each satellite as needed. Though straightforward in its im plementation, disadvantages to the traditional approach include the la rge workload placed on the ground station, the need for multiple groun d stations to view satellites in different geosynchronous positions, a nd unaccounted-for delays in atmospheric propagation. In early 1996 Mi lstar became the first satellite system to employ crosslinks for preci se satellite time synchronization. At that time, the crystal oscillato r clock onboard FLT-1, the first Milstar satellite, had its time and f requency tied (i.e., slaved) to the rubidium (Rb) atomic clock carried onboard FLT-2, the second Milstar satellite. The FLT-2 Rb atomic cloc k was controlled by the ground, while the slaving of FLT-1 to FLT-2 wa s accomplished without ground intervention: all timing information req uired by the slaving algorithm was obtained through the FLT-1 to FLT-2 satellite crosslink. In this paper we will first show the timekeeping capabilities of the two satellite clocks when operating independently , which indicate that both clocks are performing well. Then, we will p resent ground station measurements of FLT-1 and FLT-2 timekeeping that demonstrate satellite synchronization to better than 150 nsec without ground intervention. As satellites are added to the Milstar constella tion, crosslink slaving will minimize ground station timekeeping activ ities, thereby lowering system operating costs. (C) 1997 by John Wiley & Sons, Ltd.