NavList:

Who's that Nobel-honored physicist who got in wrong?

No, not inertial and not celestial, though it's certainly possible that the US GPS satellites or some of the other satellites in other GNSS constellations use inertial or celestial "cues" or corrections to enhance the primary navigation solution.

Highly accurate satellite ephemeris data are essential to ground-based position finding. While GPS satellite orbits are normally determined by the "ground segment" working the navigation problem in reverse from precisely known locations, the system is also capable of consistent inter-satellite monitoring of satellite orbits. This has been possible from an early date (more than thirty years ago by now). The satellites listen to each other's signals and by various processes, including "simple" averaging, they can can detect and eliminate drift in each other's atomic clocks and generate their own orbital ephemeris data. From my very limited knowledge on this topic, it's estimated that the satellites can function autonomously without any contact whatsoever with the ground segment for a minimum of six months. The actual period of autonomy could be somewhat longer, or it could be significantly longer. And it's classified. That's a good trick and mostly un-necessary. What would knock out the ground segment?! But that's one reason it's nice to have a system designed and maintained by the Pentagon. They think about Armaggedon.

Do other GNSS satellite constellations have this same autonomous capability? Presumably some do, and some do not. The Russian Glonass and the Chinese Beidou are both managed by their countries' respective militaries, and they, too, think about Armaggedon. Meanwhile the European Galileo constellation has been built primarily with civilian users and very high accuracy in mind. The end of the world may be a lower priority...

Frank Reed