From W. Fri Apr 23 14:26:00 1993 From: W. (W.) Date: 23 APR 1993 14:26:00 Subject: [IGSMAIL-246] RINEX under AS Message-ID: IGS Electronic Mail 23-APR-1993 14:26:00 Message Number 246 *********************************************************************** >From: W. Gurtner Subj: RINEX under AS ---------------------- RINEX under AS ______________ Werner Gurtner Astronomical Institute University of Berne The current versions (1 and 2) of the Receiver Independent Exchange Format RINEX knows the following 7 basic observation types: C1: C/A code pseudorange on L1 P1: P code pseudorange on L2 P2: P code pseudorange on L2 L1: phase observation on L1 L2: phase observation on L2 (D1: Doppler observation on L1) (D2: Doppler observation on L2) We have never distinguished between L1 phases recovered using C/A code only or using the P-code. In the presence of both only one (the better one) has been included in the RINEX files. Half-wavelength phase observations (squaring receivers) have been flagged - by the wavelength factor (set to 2) for all satellites in the header of Version 1 files - using either a general or a satellite-dependent wavelength factor in Version 2 - using bit number one (LSB being bit 0) in the loss-of-lock indicator (LLI) to mark observations with a wavelength different from the one given by the wavelength factor (Version 2). This is used for instruments that may frequently switch from p-code tracking to squaring and vice versa due to tracking reasons With the introduction of receivers that do not merely switch to C/A code tracking and sqaring under anti-spoofing (AS) conditions but use cross-correlation or other techniques instead, the question came up how RINEX will handle these new observation techniques or, strictly speaking, these new observation types. The discussions in the user community (especially within the International GPS Geodynamics Service (IGS) group) showed two different approaches: - use the actual observations to synthesize the corresponding basic RINEX observation types and flag them to be "special" - define enough new observation types to cover all the different ways the new receivers are or will be handling the AS condition Apart from the fact that some observation types might not be available at all or that phase observations are collected in half wavelength under AS, the most conspicuous result is an increased noise of the AS-affected observables. As a consequence some processing software might have to know what observables fall into this category. In order to not have to go through the process of actually creating a new RINEX version (3), a process becoming more and more difficult, the data exchange working group of IGS came to the conclusion to propose the following moderate change of RINEX version 2: Observables under AS having been synthesized using cross-correlation or other techniques and therefore suffering from an increased noise level are flagged with bit 2 of the loss-of-lock indicator! (To set bit 2 of the LLI its current value is increased by adding 4.) This will most certainly affect the L2 observations (cross-correlation, P/W tracking) P2 observations (cross-correlation, i.e. P2=C1+(Y2-Y1) P1 and P2 obs (P/W tracking) Rogue / Trimble SSE types of instruments track C/A code on L1 under AS: Their AS code observation on L1 in the RINEX file is "C1", the same one as from any C/A tracker under non-AS condition. True Y-code trackers would create unaffected P1,P2,L1,L2 observables that do not have to be flagged. Squared L2 observables of course have to be flagged with wavelength factor 2 as described above. When AS is switched on the LLI flag will go from (usually) zero to 4 (or most certainly for one epoch to 5, as the receiver will loose lock). Observations of receivers that go to squaring mode will either be additionally be flagged with bit 1 of the LLI or a new header line (WAVELENGTH FACT L1/2) might be inserted according to the RINEX version 2 rules. Software that does not care about the subtleties of AS-affected observations can ignore bit 2 of the LLI and just has to worry about possible different wavelength factors. I also got the proposal to introduce a floating point version number to be able to flag this small change with a version such as 2.1. Most of the people I contacted did not find it necessary, however, and nobody voted for a version 3. Please find below an example of a RINEX file containing satellites under AS (prn 2,27,24,26,19,16) as well as non-AS satellites (prn 3,18). I hope to help decrease or even eliminate with this step uncertainties in the handling of AS-affected data by the different RINEX converters and by the large variety of post-processing software. May I ask everybody who is providing RINEX converters to follow this proposal to help prevent the development of many different RINEX "dialects". Do not hesitate to contact me if you have any questions. Maybe it is a good idea to include a COMMENT line stating that bit # 2 of LLI flags AS data. Berne, April 23, 1993 Werner Gurtner PS: A complete RINEX version 2 format description is available as ASCII file (RINEX2.TXT) on our anonymous ftp account (Internet 130.92.4.10) in the directory [ASTRONOMY.RINEX]. ****************************** EXAMPLE ********************************** 2 OBSERVATION DATA G (GPS) RINEX VERSION / TYPE TRRINEXO V1.8.1 L+T 21-APR-93 00:38 PGM / RUN BY / DATE Zimmerwald LT88 COMMENT BIT 2 OF LLI (+4) FLAGS DATA COLLECTED UNDER "AS" CONDITION COMMENT ZIMM MARKER NAME LOGST/COMPAQ L+T OBSERVER / AGENCY 2691 TRIMBLE 4000SSE 5.60 REC # / TYPE / VERS 17498 TRIMBLE 4000SSE ANT # / TYPE 4331354.9039 567542.1797 4633078.2930 APPROX POSITION XYZ .0000 .0000 .0000 ANTENNA: DELTA H/E/N 1 1 WAVELENGTH FACT L1/2 5 P1 L1 L2 P2 C1 # / TYPES OF OBSERV 1 INTERVAL 1993 4 20 6 14 .000000 TIME OF FIRST OBS END OF HEADER 93 4 20 6 14 .0000000 0 7 2 27 3 24 18 26 19 -3606452.220 5 -2658497.69346 24290247.9384 24290242.383 -17427926.331 8 -13467468.00549 20337177.6454 20337176.836 24609815.414 1492000.824 2 2024814.857 6 24609819.145 4907925.958 5 3983864.90846 23660590.2034 23660586.875 23624383.094 11049247.283 5 8651874.527 6 23624385.047 -986672.123 3 -567492.38945 24759926.0744 24759926.203 -6169648.462 6 -4701382.96849 21418332.6724 21418331.031 93 4 20 6 14 30.0000000 0 8 2 27 3 24 18 26 19 16 -3679377.118 5 -2715322.21546 24276368.9304 24276363.586 -17416465.956 8 -13458537.83849 20339358.3524 20339357.508 24622192.492 1557030.139 3 2075487.050 6 24622194.531 5028029.695 5 4077452.15346 23683446.9884 23683443.461 23646913.781 11167645.148 5 8744132.522 6 23646915.785 -1046564.357 3 -614161.67545 24748527.3754 24748527.445 -6080119.925 6 -4631620.48549 21435369.8714 21435367.984 -9083036.05917 -6998390.56559 21002894.8014 21002893.266 93 4 20 6 15 .0000000 0 8 2 27 3 24 18 26 19 16 -3752199.504 5 -2772066.79546 24262513.3834 24262507.102 -17404473.055 8 -13449192.72049 20341640.3754 20341639.742 24634693.094 1622726.990 2 2126679.334 6 24634696.039 5148356.362 5 4171213.11046 23706343.6644 23706339.984 23669457.367 11286114.710 5 8836446.376 6 23669459.344 -1106381.016 4 -660772.02245 24737147.7424 24737145.891 -5990242.516 7 -4561586.17849 21452473.0664 21452471.180 -9096979.011 7 -7009255.21749 21000242.1174 21000240.078 ************************** END OF EXAMPLE ********************************* **************************************************************************** Werner Gurtner, Astronomical Institute, University of Berne, Sidlerstrasse 5 CH-3012 Bern, Switzerland, Tel: ++41-31-658591 / 658599, Fax: ++41-31-653869 E-Mail: GURTNER at AIUB.UNIBE.CH GURTNER at 130.92.4.10 20579::49203::GURTNER