[IGSMAIL-281] Rogue Version 7.4 Upgrade

[U.J.]

IGS Electronic Mail       10-JUL-1993 10:24:30       Message Number 281
***********************************************************************
 

>From: U.J. Lindqwister
Subj: Rogue Version 7.4 Upgrade
-------------------------------

The Rogue software upgrade (Version 7.4) does not solve all problems with
the CA pseudorange. The upgrade did, however, insure that the bias between
CA and P1 for a given receiver was the same for all satellites (when AS is on),
whereas before the upgrade the bias between the CA and P-code pseudoranges
would differ for different satellites. For example, the 10xC1-P1 difference
would be +10 meters for satellite X and -10 meters for satellite Y, etc.
Now, with the software upgrade (Version 7.4) this bias is the same for
all satellites for a particular receiver. So for the Rogue development
receiver used in the lab at JPL this bias is ~25 meters for all satellites.
For the Rogue at Yaragadee it is ~20 meters. (Note: there is one exception,
GPS19 has very low SNR for the CA band and this data appearantly has
an extra bias due to the fact that it is processed differently in the
Rogue CA filter - this is a satellite problem at the moment).

If all the satellites tracked by the Rogue were in AS mode, then with
the new upgrade this would look as a receiver clock error, which can
be removed rather easily with standard data analysis techniques.
For a while, as long as
Block I satellites are in the constellation and while some Block II's are
not encrypted (while others are), there will continue to be a 
problem with mixed biases between C1 and P1 satellite data. There
are a couple of workarounds/solutions while a better fix is being
worked out (Using Version 7.4 AS data):

1. The bias can easily be computed if you have access to the CA and
P-code data (Conan will give you only the P-code data, so you need
to upload TRK format data - it is one if the 'hidden' formats under
the DATA FORMATS and TYPES menu)
and then the CA pseudorange data can be calibrated for this bias. 

2. For those who process both pseudorange and carrier phase simultaneously
in the filter one can de-weight the pseudorange by 100 meters or more for
for AS data - not too much data noise so that the clocks can still be fixed.

3. It is also possible to solve for two clock biases for each receiver,
one for the P-code data and one for the AS data, although this would
diminish some the GPS data strength in the analysis.

The reason for the bias in the Rogue SNR-8, 800, 8C, and 8A (but not in
the TurboRogues) is because the CA data is processed through a different
analog filter compared to the P-code data filter. Since, the filters
are analog parts, the computed bias between C1 and P1 will gradually
change in time as the hardware degrades and also be different for
different Rogue receivers. The degradation is very slow, currently
estimated at 0.1%/year, but it will still require a calibration every
now and then, i.e, 10xC1-P1 must be re-computed periodically.

The current plan is to perform the AS bias calibration in the Rogue
and hence remove it before the user sees it. This new upgrade
to Version 7.5 will probably occur in late August of 1993.

- Ulf

=======================================================================
Ulf J. Lindqwister                 Telephone:   (818) 354 1734
GPS Networks & Operations Group    FAX:         (818) 393 4965
Jet Propulsion Laboratory          Email:       ujl at logos.jpl.nasa.gov
4800 Oak Grove Dr.
Pasadena, CA 91109
=======================================================================