[GRLUG] OT, but GPS, anyone?

[Michael Mol]

On Wed, Aug 26, 2009 at 12:45 PM, L. V. Lammert<lvl at omnitec.net> wrote:
> At 12:28 PM 8/26/2009 -0400, you wrote:
>>But if that doesn't do it for you then I suppose the absolute minimum
>>spacing is the GPS resolution which is normally 1 meter. Anything under
>>that and gps positional uncertainty will overtake your sample accuracy.
>>It would be best to space at least twice the GPS resolution but the
>>farther apart the two points are the better your accuracy.
>
> BUT taking a differential reading should theoretically cancel any position
> innaccuracy over a statistical sampling period, assuming identical GPS
> units and identical satellite usage - which is why I'm looking for any
> resources that might be available to answer the question.

The accuracy of your differential is going to improve with the
distance of your units, but as your units get farther apart, they're
going to see different satellites at different times, and one may find
themselves in RF shadow conditions not identical to that of the other.

Directionality may be a problem, too, if your core unit winds up
shielding one of your units from view of a satellite that the other
has a fix on.  In short, don't depend on identical satellite
reception, let the GPS units abstract that figure away individually,
unless you're planning on processing the GPS signal yourself.

-- 
:wq