RTKPLOT does not show multipath for Galileo and Beidou from ZED-X20P data

[alsper78]

Hi.
With the RTKCONV in b34L version I can get rinex observations data for GPS, Galileo and Beidou satellites from ZED-X20P raw data. But when I open the result with RTKPLOT I see multipath only for GPS, not for Galileo and Beidou.
When I work with data from ZED-F9P I don't have this problem.
X20P uses Galileo and Beidou satellite frequencies which do not correspond to GPS L2 frequency but to L5. Maybe is multipath calculation only possible with both L1 and L2? Is this not possible with L1 and L5?

[JensReimann]

Yes, I can confirm, that at the moment multipath is only calculated using L1 & L2.

Making the code more flexible to also use the combination L1/L5 or L2/L5 shouldn't be too hard (see plotdata.cpp:updateMp()).

[rtklibexplorer]

I agree and hope to get mulitpath plots for the additional frequencies into the next release

[ourairquality]

Btw if people are getting data from the X20 then it would be useful to know the actual signals received on BDS2 and BDS3? The specs appear to show no 7I (or 7D) on BDS3, and does it measure 2I and 5P on BDS2?

Knowing what to target would help in developing patches to better support it.

Would also be curious if people have tested the MB2 antenna? Initial results that I have seen suggest it picks up all the signals and that a fix can be obtained readily (better receiver than the X20), but there are hints that the PCO/PCV might be rather large perhaps around 10mm as rotating this antenna can move the result around 10mm horizontally depending on the satellite positions?

[ourairquality]

Sorry, expect the X20P BDS2 to be B1I, B3I [2I 6I] not B2a [5P] which appears to be BDS3 specific, and these also appear to be the signals used for the IGS products for the reference clock.

[alsper78]

I agree and hope to get mulitpath plots for the additional frequencies into the next release

That would be awesome. Thank you.

If it can be useful to you, these are the satellites shown by RTKPLOT in the L5, L6 and L7 bands (a test of about 15 minutes with ANN-MB2 antenna):

Here (https://drive.google.com/file/d/1EL-_3wx20xNHX69YVKZiXx6RDGXvcr0t/view?usp=sharing) the log file with UBX_RXM_MEASX ,...RAWX, ...SFRBX, NMEA_GNGNS, ...GNRMC, ...GNVTG data.

[ourairquality]

Thank you for sharing this data. Issues:

• The BDS B1 [2I] appears to be missing in the data, it not there for BDS2 or BDS3. BDS2 decodes to only B3 [6I] (sigid 4). BDS3 decodes to B1C [1P] (sigid 5), [5P] (sigid 7), and [6I] (sigid 4) which is a useful set but might have also included B1 [2I].

• The BDS3 B1C [1P] signal currently maps to the rtklib frequency index 4 so the current default build with NFREQ=4 will not be able to process this in a solution (but can in data processing).

There did not appear to be any errors produced by the decoder, and there did not appear to be any unmapped signal ids for BDS.

The other systems, GPS GAL SBAS, appeared to decode the expected signals.

A quick search found the following interface description for the X20P but it did not help.
https://content.u-blox.com/sites/default/files/documents/u-blox-20-HPG-2.00_InterfaceDescription_UBXDOC-304424225-19888.pdf

A suggestion would be to try ublox ucenter and see it it reports the signals being received and if so then check if BDS B1 [2I] is reported, and also to log some RTCM3 data and see which signals are reported.

There are issues in RTKLIB to work on. I have a wip patch to separate BDS2 and BDS3 so they can have separate rtklib frequency index assignments, and so that they can be processed as separate systems to deal with the bias between them. Still considering how best to make the frequency index assignments a runtime config option, but this is perhaps easier with some other patches in the pipeline such as centralising the frequency index mapping. Also how this works with specifying the IFLC.

[JensReimann]

I have created a pull-request to solve the multipath issue (see #707).

Can please someone check the win app version and confirm that this patch solve the issue.

[alsper78]

About the frequencies, the X20P integration manual (https://content.u-blox.com/sites/default/files/documents/ZED-X20P_IntegrationManual_UBXDOC-963802114-12901.pdf) at pag. 14 says the default configuration gets this:

• GPS: L1C/A, L2C, L5
• Galileo: E1B/C, E5a, E6
• BeiDou: B1C, B2a, B3I
• QZSS: L1C/A
• SBAS: L1C/A

In u-center-2 I see this:

Here (https://drive.google.com/file/d/1Mq6qoPp7LxgiOLJTpzX1bS1Iop_sQCBz/view?usp=sharing) is a UBX file recorded by u-center-2 with also the RTCM data as suggested by the interface manual. pag. 24:

This test was made with a Budget Survey Tripleband GNSS Antenna (https://www.ardusimple.com/product/budget-survey-tripleband-gnss-antenna-ip66/)

[ourairquality]

Thank you, that helped. Found these config messages in the interface description at '6.9.22 CFG-SIGNAL: Satellite systems (GNSS) signal configuration' and the defaults on page 218, but did not find a full list of supported combinations. The question remains if CFG-SIGNAL-BDS_B2_ENA can also be enabled to also receive BDS 2I, or if this can only be enabled if perhaps CFG-SIGNAL-BDS_B2A_ENA is disabled, as did not find a table of supported combinations for BDS. Similarly for Galileo E5B.

The data is as expected for the default config, with BDS 2I disabled, it reports B1C and B2a which are enabled by default and the antenna does not appear to support the L6 band so that block B3I and also Galileo E6. There are no BDS-2 satellites in the above ucenter signal plot but this is expected as the only signal enabled for them is 6I which the antenna does not appear to support.

For this antenna it might be possible and useful to disable all the L6 band signals in the config and then try enabling GAL E5B and BDS B2. There is also CFG-SIGNAL-PLAN, but what is the difference between Plan-1 and Plan-2?

Fwiw the ANN-MB2 antenna did pick up all the signals including the L6 band and the price seemed competitive for an all-band antenna and it is more compact than budget survey antennas such as the Beitian BT300S (which does receiver the L6 band).

We'll need to get rtklib working well with these bands and signals.

[alsper78]

Sorry, I noticed that u-center-2 by default didn't show all satellites and frequencies. Now I have it set to show everything, even satellites not used for navigation.
This is the result.

With the MB2 antenna:

With the budget-survey antenna:

The results are quite similar. Both the antennas appear to support L6 band.