[MLB-WIRELESS] OT: increasing GPS accuracy

Tim Hogard thogard at knotty.abnormal.com
Tue Oct 17 13:10:25 EST 2006

> It's not the cost of the DGPS receiver, they were relatively cheap ( 
> say $200 ) but you need to have a DGPS base within range.

Plans for one are here:

The problem with WAAS in Australia is that Air Services has killed
it.  They want to do their own thing that will only work for airplanes
and will be useless for boats or other GPS users.  The USAF has
stated that they will transmit Aussie WAAS info if its provided to
them.  Japan will also but want to get paid for it.

I didn't make it to the CGSIC meeting in Brisbane where this was
discussed.  Its a political problem that will require someone to
spend a lot of time fighting Air Services and CASA.

There is alos another group (that had a meeting yesterday?) that
is building referenc stations in Victoria.

The rest of this wasn't sent to the list so I'll tack it on here:
> Just a quick OT question as I know many of you fiddle with GPS: if you had
> a reference GPS in a fixed location with long-term averaged co-ordinates,
> could you use the difference between the current and long-term reading
> from that GPS, to correct another (portable) GPS unit nearby?
> eg, say the long term position is x,y. At 09:01:30 the current reading is
> x+3,y-2. Can this offset be retroactively applied to the reading saved at
> the same time from a portable GPS?

Yes and there are several ways.  What your talking about is sometiems
called "poor man's dgps" and it only works if both recievers are
using the same set of sats.  A better way is called post processing
and it works very well but its complicated.  To make it work, you
need pseudorange from the GPS reciver which tells you that sat 3
is x meters away at time t.  Then you run that and the diferential
through the same forumlas that a GPS uses internally and you cna
get a very accurate fix.

A good modern GPS will give you range and phase which is enough
for sub cm precision.

Some hints on how to do it are on Wolfgang's site:


More information about the Melbwireless mailing list