Re: gpsd ignores Galileo satellites

[Filip Jan Kubicz]

Hi Gary,

>> Thank you for checking. Let's stick to the first log I have sent (and
>> ignore the ones with "ANR01A02SV03" in the filename - they're not
>> correct).
>
> Too late, already in git head.  What is incorrect?  Beyond that they are
> all incorrect?

The firmware "ANR01A02SV03" is some debug firmware and it gives output incompatible with the documentation. Output gathered from the factory firmware is compatible with the doc.

>> I have analyzed the GSV and GSA messages and they are OK according to
>> the Quectel description.
>
> Which does not make them valid NMEA.  Take this:
>
> $GNGSA,A,3,103,132,,,,,,,,,,,1.6,0.9,1.3,3*3D
>
> Satellite IDs must be from 1 to 99.  What does 103 mean???  Broken.

Yes, looks like the Quectel format is not compatible with NMEA0183 despite the claims.
I'm waiting for more documentation from the vendor. If this makes sense, I could add the support for this data format to gpsd parser.

Kind regards,
Filip

On Wed, Dec 2, 2020 at 6:19 PM Gary E. Miller <gem@rellim.com> wrote:

Yo Filip!

On Wed, 2 Dec 2020 16:48:32 +0100
Filip Jan Kubicz <filip.kubicz@tier.app> wrote:

> Thank you for checking. Let's stick to the first log I have sent (and
> ignore the ones with "ANR01A02SV03" in the filename - they're not
> correct).

Too late, already in git head.  What is incorrect?  Beyond that they are
all incorrect?

> I have analyzed the GSV and GSA messages and they are OK according to
> the Quectel description.

Which does not make them valid NMEA.  Take this:

$GNGSA,A,3,103,132,,,,,,,,,,,1.6,0.9,1.3,3*3D

Satellite IDs must be from 1 to 99.  What does 103 mean???  Broken.

> $GPGSV,3,2,07,22,73,081,22,1*5F
> $GPGSV,3,3,07,01,57,157,23,09,11,221,27,8*66
> $GLGSV,1,1,03,70,17,250,30,83,18,173,27,68,,,20,1*70
> $GAGSV,2,1,04,103,23,231,37,136,34,164,34,6*75
> $GAGSV,2,2,04,103,23,231,26,136,34,164,09,1*7F
> # breakdown:
> 04 Galileo satellite signals in total
> PRN 103, elev 23, azim 231, CN0 37
> PRN 136, elev 34, azim 164, CN0 34
> - signal ID for above: 6 (Galileo E1)
> - the same 2 satellites send signal ID: 1 (Galileo E5a)

Pre NMEA 4.11: there is no PRN in GLGSV or GAGSV, there is only
satellite numbers from 1 to 99.  Even if it was a PRN, there is no valid
PRN 103 in NMEA.  PRN 136 is SBAS, not Galileo, so also invalid.

> $GNGSA,A,3,70,83,,,,,,,,,,,3.5,2.2,2.7,2*3D
> # Breakdown
> $
> GN - Talker ID (multiconstellation)
> GSA,
> A - mode: Automatic switch between 2D/3D
> 3 - 3D fix
> 70, 83 - Satellite IDs
> 3.5 - PDOP
> 2.2 - HDOP
> 2.7 - VDOP
> 2 - SystemID (2 = GLONASS)
> *3D - checksum

gpsd does not handle SystemID in GSA.  Your sample is the first we
have seen.  Dunno what to do with it.  To start gpsd will just
ignore GNGSA.

> If we ignore the satellite IDs, which are not exactly mapped to PRNs,

Not "exactly"?  How about 100% undefined and illegal.

> then the GSA is compatible with GSV. Looking at the above breakdown,
> do you have an idea which part is not handled by gpsd's NMEA0183
> parser and causes the Galileo satellites to be omitted?

Invalid, and undocumented, satellite IDs.  gpsd is correctly ignoring
invalid data.  gpsd does fix bad output from some receivers, but that
is only possible with documentation.  Quectell let me see their "NMEA"
doc, it does not explain their "PRN"s.

I have a request in to Quectel to explain their "Querky" take on NMEA.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        gem@rellim.com  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin