[OSM-talk] Cooperative Differential GPS

[Tapio Sokura]

Gora Mohanty wrote:
> are largely due to atmospheric fluctuations. Corrections for these are
> derived from the difference between the  well-known actual position of the
> base station, and the received realtime position. These can be broadcast
> through various means, e.g., an Internet DGPS server, and should be good for
> a few hundred km around the base station. The cooperative aspect of this
> comes in because of the benefits of multiple base stations.

Proper differential corrections for GPS are about measuring and 
correcting pseudoranges, the measured distances between GPS satellites 
and the receiver. SBAS systems (WAAS/EGNOS/MSAS/etc) and RTCM SC-104 
(DGPS) do the corrections differently, but the basic idea is the same, 
to correct the pseudorange measured by the GPS receiver. There are also 
other mechanisms to doing differential corrections, but they are mostly 
used in the surveying sector, either real-time or post-processed.

Anyway SBAS/DGPS corrections can't be applied to a computed position 
(lat/lon/alt). The corrections also can't be generated at a reference 
station by just calculating the difference between a position returned 
by uncorrected GPS and the real/surveyed position. This is because you 
can't guarantee that two receivers, no matter how close together, use 
the same satellites and identical weighting and filtering in the 
position solution.

So you need to go down to individual satellite pseudoranges. SBAS 
systems actually go down even further than that and break the 
corrections into error components like satellite clock error, orbit 
error and ionospheric delay. This means that both at the correction 
generating end and at the correction use end the GPS receivers have to 
have support for differential GPS built-in. Or as an alternative, you 
have to have access to the pseudorange measurements and have software 
that applies/generates the corrections and calculates the positions. You 
just can't post-correct a typical NMEA position ($GPRMC/GGA) and expect 
to get good and reproducible results.

The Internet DGPS systems based on relaying RTCM SC-104 corrections work 
ok, if you are within a couple of hundred kilometers from the reference 
station. Multiple correction streams can't easily be combined, you have 
to somehow select one (typically the one geographically closest to you).

SBAS systems supply corrections to a large geographical area at once, up 
to a continent (or two). But the trouble there is how to get the SBAS 
corrections into your GPS receiver, if you aren't receiving the 
corrections directly from the geostationary satellites. For example the 
EGNOS correction stream is available via Internet from SISNeT, but a 
typical GPS receiver can't handle the corrections input via it's serial 
or USB port. At least in theory you could convert the SBAS data into 
virtual RTCM SC-104 reference stations and feed that into receivers that 
support it. But then again nowadays RTCM SC-104 support often just isn't 
there in the GPS receiver firmware..

   Tapio