[Xastir-dev] Position accuracy

[Gerry Creager (N5JXS)]

Curt Mills, WE7U wrote:
> On Sun, 7 Dec 2003, Gerry Creager N5JXS wrote:
> 
> 
>>The compressed positions with resolution of 1 foot are really
>>misrepresentations.  Bob B.'s said this, and I've actually agreed with
>>him, although not always for the same reasons he stated.
> 
> 
> Care to elaborate?  What is the resolution (roughly) that one can
> expect from the compressed protocol?  I know it's much better than
> the normal APRS position resolution.

Well, unless you're using Real Time Kinematic hardware, you're likely 
seeing 6m data at best.  If we do a careful analysis of the GPS signal 
and system, you should not expect anything better than somewhere between 
29-45m from a conservative||pessimistic evaluation of the error budget 
for well designed receivers with adequate processors and code, using 
code-phase determination of autonomous position on L1 C/A codes.  This 
ignores augmentation by conventional DGPS, or WAAS.  These should offer 
an improvement of UP TO an order of magnitude, but may actually degrade 
accuracy if you're not careful...  I can get into the mechanics of this 
later if you like.

NOW:  Is that 0.1 x 29m or 0.1 x 6m?  And how'd we get 6m anyway?  Well, 
we got it empirically, and that started literally the minute we saw SA 
turned off.  Most of the time, well over 80%, usually >95% (for suitable 
intervals) we see accuracies on the order of 6 minutes, that are 
systemmatic and pseudorandom or random.  Augmentation will improve the 
REAL state of accuracy, in my experience, by a factor of 10.  The error 
budget, however, of the augmentation mechanism must also be evaluated. 
And there's the rub.  For poor selection of WAAS platforms, you can in 
fact degrade performance.  Similarly, I've seen this happen with RTK 
installations, and regular DGPS sites.  Typically, the user's fat dumb 
and happy, because their receiver is claiming some incredible error 
figure, that has no basis in reality.

>>When you're entering data for a position, it's considered good practice
>>in the geodetic surveying community to enter the coordinates as
>>precisely and accurately as possible.  That way, if there's a subsequent
>>loss of precision, it's not for want of good initial data.
> 
> 
> Kind of like getting the best possible S/N ratio at the initial
> injection point of the signal into the receiver...  Garbage in,
> Garbage out.

Exactly the point.  We should not be sending a position represented to 
more precision than we have data to support!

73, gerry
-- 
Gerry Creager -- gerry.creager at tamu.edu
Network Engineering -- AATLT, Texas A&M University
Office: 979.458.4020  FAX: 979.847.8578
Cell: 979.229.5301    Pager: 979.228.0173