Re: GPSD For Real Time Application

[Greg Troxel]

In addition to Gary's advice, I'll point out that the notion of accuracy
and high speeds is complicated, as it gets into the timestamp of the fix
and where the receiver (antenna) was at that time.  The next question is
if you interpolate to get positions in between fixes.

If your 5m is cross-track error, that's another story.

Also, you said "accuracy required is about 5m", but you didn't explain
if that was CEP, 1 or 2 standard deviations, 95th percentile, or 99th
percentile.  See gpsprof and the man page, and this document:

https://www.wsmr.army.mil/RCCsite/Documents/261-00_Standard%20Report%20Format%20for%20GPS%20Receivers%20and%20Systems%20Accuracy%20Tests%20and%20Evaluations/261-00.pdf

and then perhaps ask a clearer question.  Include in the question your
assumptions about the antenna and the sky view/horizon obstructions.
Include your results from running gpsprof with the vehicle static for
24h, and explain what you are doing about choosing which constellations
you are using and what your plan is in terms of SBAS or other
differential corrections source.

If what I said is significantly confusing, you probably need a
consultant :-)