I understand it’s more computationally intensive then not encrypting or symmetrical encryption. But we have better and better processors in low power form-factors. Is this so computationally intensive that it couldn’t be done my an iPhone-class CPU? A Mobile i-class CPU like laptops and NUCs use? How intensive is it?
I understand that this would raise the cost of GPS (again), but GPS is pretty darn cheap these days. I get that more processing means less battery, but if we are dealing with passenger craft, the differences would seem to be minimal. Container ships, cruise ships, yachts, airliners all could benefit.
And for our cellphones, cars and drones, we could keep the old system, understanding that it’s not foolproof.
Unless we’re dealing with server rack computational needs, this seems to be a problem of cost and will, not capacity.
But it is also entirely unnecessary. You don’t need to sign the entire stream. You just need to sign the occasional message. If all the GPS satellites send out signed updates at around the same time, the receiver will be able to calculate an authenticated position and any data from outside that known position must be spoofed can can be ignored.
I suppose you could also have a receiver with an exceptionally good clock and 3 antennas. You could then triangulate the position of any signal source and see if it is coming from space where the satellite is supposed to be.
My favourite thing about GPS is how it demonstrates the reality of relativity to people* who prefer to file it under “crazy things that scientists think up but can’t prove”.
Clocks higher in a gravity well (like those in GPS satellites) run faster than clocks lower in that well (like those in iPhones). GPS signals are timestamped, and these timestamps are critical to the calculation of a GPS receiver’s position. So if the satellites clocks are “wrong” (from the receiver’s perspective), so will be the calculated position, Hence why the clocks in GPS satellites are deliberately slowed, as laid down in the GPS technical specs (pp. 39-40).
QED. No Einstein, no Pokémon Go.
* I have a highly educated friend (STEM PhD, though not physics) who is very uncomfortable with relativity, sometimes to the point of rejecting it completely: he prefers his universes Newtonian. This is my favourite stick for beating him with. I’m still working on my “no quantum mechanics = no electronics” pitch for him.
