[time-nuts] What position is measured?

[jimlux]

Tom Holmes wrote:
> One other delay contributor  would appear to be processing delay in the
> receiver, which thus begs the question of how the PPS signal is actually
> synchronized to the GPS system.
> 

The GPS nav messages is synchronized to the seconds, so it's a matter of 
making sure the output pulse is synced to the appropriate time in the 
GPS signal.  The delay in the receiver is (reasonably) constant, so the 
mfr essentially calibrates it out.

It's not done precisely like this, but conceptually, you have a 1pps on 
the spacecraft driven by a Cs clock, you receive the signal in your 
receiver (some time later than the actual "change of second") and 
subtract out the light time delay from satellite to you. (or, more 
accurately, delay the signal from the receiver to the "next" second).

It's controlling for that "light time delay" that's the tricky part, 
since it varies depending on the degree of ionization of the ionosphere.