[Xastir] Question about APRS GPS position precision

[Steve Friis]

Andrew Rich wrote:
> Yeah but in the real world, it still puts me off the road or runway.
>
>
> ----------------------------------------------------------------------------
> Andrew Rich VK4TEC
> vk4tec at people.net.au <mailto:vk4tec at people.net.au>
> http://www.tech-software.net
>
>
>
>
> -----Original Message-----
> From: xastir-bounces at xastir.org [mailto:xastir-bounces at xastir.org]On
> Behalf Of Gerry Creager
> Sent: Monday, 8 October 2007 6:54 AM
> To: Richard Polivka, N6NKO
> Cc: Jim Tolbert; XASTIR
> Subject: Re: [Xastir] Question about APRS GPS position precision
>
>
> Richard Polivka, N6NKO wrote:
>   
>> Jim,
>>
>> Most GPS units are good to 4 decimals. Any higher precision requires
>> post-processing or L1/L2 reception (not avail in consumer equipment).
>> When you factor in multipath and all the other variables, 4 decimals is
>> quite good but it takes time and patience - think searching for a
>> geocache in a forest. Plus, at four decimals, on a patch antenna minus
>> ground plane, it is quite unstable.
>>     
>
> 9 cm more or less should be plenty good enough for most of our users.
> That's 4 decimal-place precision.  That said, an L1 signal (L5 won't be
> available for some time still) position assuming really good geometry
> and a stable antenna platform is likely to be good only to ~6m
> horizontal and ~13.7m vertical... at best.
>
>   
>> Plus, I have a feeling that when Bob B. designed APRS, he was not
>> looking at this being used for what we are doing.
>>
>> Until the data output is smoother and better accuracy, five decimals in
>> - broadcast 4 - rewrite the standard, this may be the best for now.
>>     
>
> The limitations in precision are in rank order, the spec and the spec.
> For accuracy the limitations are:
> User equipment antenna configuration
> Ionosphere
> Troposphere
> Multipath
> GPS Signal Specification for L1
>
> When I resolve cm accuracy, or better, I do it using dual-frequency
> (L1/L2) receivers, multiple stable baseline processing on ground-plane
> or choke-ring antennas, at a fixed and measured height about the ground,
> and post-process the data to include a least-squares adjustment of the
> position.  The process is as much statistical as matrix-mathematical in
> accomplishment.
>
> gerry
>   
Well, worst case is I am still within shouting distance. Not bad by any 
means.

Steve/WM5Z