Poor Man's DGPS> Why you need not bother.

Method #1
"I want to  leave one (ordinary consumer type) GPS receiver at home  at a fixed position and recording position and take the other identical GPS receiver  out to the field and record position.  Then,  when I get back home,  I will compare the two recorded position outputs from the two receivers and subtract out the SA and other errors. "

QUESTION:

HOW ACCURATE WILL THIS BE?

ANSWER:

This "poor man's DGPS" approach will not give you any improvement you can DEPEND ON.  At some moments,  it might be accurate and sometimes it won't and you will never be able to tell WHEN it is accurate and when it is not.  The reasons for this are:

1)  It would be a remarkable coincidence if two unsynchronized receivers happened to choose the same satellites at the same time. The typical receiver switches among satellites fairly frequently based on proprietary criteria that vary from one manufacturer to another, one model of unit to another, and probably even one software release to another. The criteria probably include some measures of  Signal to Noise Ratio (SNR),  (or CNR) and somebody's idea of preferred geometry, and  horizon mask angle among other things.  These will vary from one location to another, even over distances of the order of a half wavelength.    L1 is around 20 cm, so separations of a half foot or more will produce different multipath conditions, hence differing SNR.

2) It is doubtful that many people have an accurately surveyed benchmark to reference their measurements against EVEN IF the problems of #1 above did not exist.

Method #2

Question:
Well,  If I can't use the "Poor Man's DGPS" as above,  can I at least log down my NMEA output from my GPS and post process the data against regularly published error data?

Answer:

No.  That will not work either.  There is no pseudo range data output from a normal consumer type GPS receiver.  On top of that,  the NMEA data tells you nothing about which satellites you were listening to when you were computing your position.  As a result,  you cannot post process the data coming out of your handy "consumer" GPS receiver.

If you NEED to have great position accuracy,  the best/easiest/cheapest thing to do is to GET real  DGPS or use a post processing survey grade system.   Some specialized receivers such as the Motorola OnCore do produce pseudo range output data streams for those who wish to proceed with the 'do it yourself' approach.

Method #3

Question:
Well,  then SURELY I can log DGPS data and then use that to correct my NMEA output data at a later time?

Answer:
No.  That will not work either.  Again,  there is no pseudo range data output from a normal consumer type GPS receiver.   On top of that,  the NMEA data tells you nothing about which satellites you were listening to when you were computing your position.  As a result,  you cannot post process the data coming out of your handy "consumer" GPS receiver.


Question:
Then what can I do inexpensively if I really want to post process my GPS data for greater accuracy?

Answer:
If you NEED to have great position accuracy,  the best/easiest/cheapest thing to do is to GET real  DGPS or use a post processing survey grade system.   Some specialized receivers such as the Motorola OnCore do produce pseudo range output data streams for those who wish to proceed with the 'do it yourself' approach.  Delorme in late 1999 came to market with a special program which is able to download pseudorange data from their EarthMate GPS receivers.  With this system,  it is possible to post process and get improved accuracy typically in the 10 to 20 meter error range.    See David Wilson's test data  HERE.

Joe Mehaffey