[IGSMAIL-69] Rogue Receivers and AS on

[Werner]

***********************************************************************
IGS Electronic Mail       14-AUG-1992 17:46:35       Message Number 69
***********************************************************************
 

Rogue Receivers and AS on
-------------------------

Werner Gurtner
Astronomical Institute
University of Berne

There have now been two weekends with Anti-Spoofing set to ON on about
10 block II satellites. As there are obviously some misunderstandings
or incorrect informations flowing around, I would like to clear a few
things up (and hopefully not increase the confusion).

The rinex data files flowing through the IGS network and coming to
our processing center for the weekends in question showed the
following properties:

- Rogue files created by DMD_TO_RINEX (JPL) did not contain any observations
  nor navigation messages of the AS-contaminated satellites!

- Rogue files created by CMP_RNX (EMR) however contained CA code, L1 and L2,
  and P2 (or better: CA1+P2-P1) observations.

- The same holds for files created by RINGO (NGS). The missing
  P1 observations for AS satellites and the missing CA1 observations for
  non-AS satellites have been replaced by a place holder -9999.000
  (which does not follow the rinex standards).

- Rinex files of other instrument types (Minimac, Trimble, Ashtech)
  contained also navigation messages of the AS-satellites as on every
  "normal" day, i.e. about 24 messages per day for block I satellites
  and 12 to 13 messages per day for all block II satellites.

  The suspicion that the AS-satellites did not broadcast any messages
  at all as it had been expressed in a recent IGSMAIL is therefore wrong.

- The P-Code Ashtech receivers do not automatically switch to squaring
  if AS is set on. Therefore the permanently (and unattendedly) 
  operating receiver in Zimmerwald did not observe L2 on the AS satellites.

The L2 phase observations of the Rogue receivers show a tremendously large
noise level, much too large to be preprocessed in the usual way.
The noise level seems to be several centimeters if not decimeters.

We believe that the data can only be processed if

- we accept the receiver-generated cycle slip flags as to be true
  and blindly introduce new ambiguities (bias parameters) for each flag

- use a very coarse detection level (several cycles) on epochs that
  don't show any receiver-generated flags.

The available rinex files however don't seem to reliably contain the
receiver-generated loss-of-lock flags. Raw files downloaded from
NGS, run through our own rinex converter, and subsequently processed
in the above-mentioned way suggest that solutions of an acceptable accuracy 
level might be possible.


Consequences:
------------

1. We are urgently waiting for a statement by the JPL colleagues
   containing

   - an assessment of the current situation
   - corrections of the above statements if they too contain
     incorrect informations
   - an outlook what is planned for the case that AS will be switched
     on again intermittently or constantly
   - guide-lines for how to best process the Rogue L2 data with
     the current quality 
 
2. We need

   rinex converters that correctly convert the AS data and pass on
   the receiver-generated loss-of-lock flags

By mere extrapolation we may expect that AS will be switched on again
tomorrow and on subsequent weekends, too!