How/why time-to-aquire...
Don Russell
first position. I suppose this would be when it is first turned on etc.
Why does it take so long? Is it just a matter of the processor speed and the
complexity of the calculations?
Thanks,
--
---
Don Russell
Dave Martindale
>I've read that a GPS unit may take up to 15 minutes to report an accurate,
>first position. I suppose this would be when it is first turned on etc.
>Why does it take so long? Is it just a matter of the processor speed and the
>complexity of the calculations?
No, it's the number of possibilities that need to be searched.
In order to receive the signal from just one satellite, the GPS receiver
needs to have a correlator operating in lock-step with a particular
sequence of bits transmitted by that particular satellite. IF the
satellite signal is actually reaching the receiver's antenna, and IF
the correlator is aligned to exactly the correct time offset (to within
a fraction of a microsecond), and IF the correlator clock rate has been
adjusted closely enough to match the satellite's Doppler shift and the
receiver's own motion, THEN you get a signal out of the correlator.
But get any single one of those parameters more than slightly wrong,
and you get *nothing* but noise - and no indication which of the various
things were wrong.
If a receiver starts up "warm", it already knows what time it is (internal
clock) and it knows roughly where it is (last saved position) and it
knows where the satellites are in the sky (almanac). Thus, it can predict
which satellites are overhead and thus worth looking for and what their
current Doppler shifts are. Thus, the receiver only has to search all
possible code offsets (about 1000 of them) and a small range of Doppler
shifts (to account for receiver motion) and it will lock to the signals.
If the receiver starts "cold" without those 3 pieces of information,
it has to search the sky for all satellites, at all possible Doppler
shifts, as well as all possible code timings. This takes a while.
Dave