Comment by Majromax
1 month ago
It sounds like Starlink uses GPS to localize the receiver, rather than for any active step in the communication link. Since most receivers are static, I wonder if an effective workaround to this is for the receiver to just remember its last GPS fix for longer, or worst-case allow a manual location specification in lieu of a GPS fix.
A user provided location cannot be trusted for geofencing purposes. A GNSS (GPS or other) is needed sooner or later. This is a legal requirement for sanction and regulation enforcement (US, if not others).
The user-provided location would only be used for the initial bootstrapping. After it connects, the Starlink network itself will localize the receiver to within 1 km.
If the user inputs a bogus lat/lon, it would simply fail to connect. There's no way to 'spoof' your location on this type of global satellite comm network.
EDIT It will be interesting to see what anti-censorship and anti-DOS hardening features are coming in future software updates. Full GPS denial bootstrapping is the most obvious, and actually this should be possible without needing to input a location. Adding offline update packages, so signed anti-denial firmware updates can "sneakernet" across oppressive regimes to recover DOSed terminals, would be even better.
Starlink system inevitably knows the terminal location down to a service cell, which is what, a 20km grain? Good enough for "regulation enforcement".
The satellites know where they are TX beamforming to a fine-enough degree of specificity for geofencing.
Assuming the jamming is coming from close to the horizon, I wonder if one could improvise a choke ring around the antenna with some sheet metal or foil.