From: Al Bowers <bowers@orville.dfrc.nasa.gov>
Newsgroups: rec.aviation.military
Subject: Re: First fighter aircraft to use INS?
Date: 08 Feb 2000 15:02:39 -0800

Mary Shafer <shafer@rigel.dfrc.nasa.gov> writes:

> Still, the el cheapo civilian handheld GPS is within ten or twenty
> feet and the military version is even better.  Differential GPS, with
> the civilian systems, is accurate to a foot or two, I'm told.  We put
> a GPS into the SR-71 when we were doing the shock field measurements
> and, once we got the software limits adjusted to keep the program from
> discarding the values determined at our cruise Mach and altitude as
> obvious errors, found the data to be astounding.  And the GPS is a
> practical system at a reasonable price for every airplane unless the
> satellites go off-line.

Actually, it takes pretty heroic measures to get accurate GPS data.
The "dither" is about 100m maximum (though three sigma is about 60m)
for raw uncorrected civilian GPS.

Differential can drop this to about 5m accuracy, which is rather
astounding.  If you start to count the individual wavelengths of the
frequency (L-band) for carrier-phase differential, you can get down to
about 4 inches.  And if you can resolve better than one wavelength
(1/8th wave is about the best I've seen demonstrated in a lab) it's
about 1cm accuracy.  It's not very fast to calculate though (typical
latency is about 1200 milliseconds in carrier-phase differential).
But the timing of the data is very good (we couldn't spoof the
one-per-second data even at very high frequencies).

We used it to measure the length of a tow rope in one experiment.
Neat stuff...

Al Bowers

--
Al Bowers    Aerodynamics    NASA Dryden Flight Research Center
"...tactics always degrade strategy..."  -Frank Bethwaite