Xiaomi Mi 9 - Incorrect, unlabeled GNSS frequencies

[TaaviE]

Summary:

A few GNSS frequencies don't have a nice name.

Steps to reproduce:

Open GPSTest on a Mi 9

Expected behavior:

What did you expect the app to do?

Observed behavior:

The nice names aren't displayed for these frequencies:

• GALILEO 1602.000 MHz , ID 7, 19
• GALILEO 1561.098 MHz, ID 7, 19, 26, 31, 31
• BEIDOU 1602.000 MHz, ID 2, 5, 9, 13, 16
• Galileo SBAS 1602.000 MHz, ID 126, 136

Device and Android version:

Mi 9, 9

[barbeau]

@TaaviE Thanks for reporting this! Could you include a screenshot too?

Unfortunately it looks like these are bugs - as far as I can tell there aren't any real signals being broadcast at these frequencies. Here's an example GNSS frequency chart:

Could you please report this bug to Xiaomi and see what they say?

[TaaviE]

I asked Xiaomi to fix these.

Just mentioning that this chart seems more complete: https://gssc.esa.int/navipedia//images/1/1c/Galileo_Signals_in_Space.png

[PedroVenancio]

Hi @barbeau and @TaaviE

I also see several strange things in my Xiaomi Mi9, with Android 9 PKQ1.181121.001 and MIUI Global 10.2.13 (GPSTest v3.4.0 and v3.4.2).

1) Several GPS satellites repeated: ID 24 and ID 30 on L1;

2) GLONASS L5 label (is this supported?);

3) Galileu frequencies of 1561.098 MHz;

4) BeiDou frequencies of 1602.000 MHz;

5) Galileu SBAS ID 136, with frequencies of 1561.098 MHz.

By the way, is SBAS supported by Mi 9?

I must say that Mi 9 is a very disappointing device to me in terms of GNSS. It is a huge drawback when compared to Mi 8... The unsupported accumulated delta range data is the worst thing...

@TaaviE What channels do you use to communicate with Xiaomi?

Thanks!

[TaaviE]

Several GPS satellites repeated: ID 24 and ID 30 on L1; Galileu frequencies of 1561.098 MHz; BeiDou frequencies of 1602.000 MHz;

Some satellites do send two signals on two bands, my hypothesis is that the reason we see same frequency twice is because I think they are reusing the same frequency lookup table (the QC chipset probably sends some byte that specifies a fixed frequency) for both bands (which is obviously wrong) and as a result we see duplicate frequencies and after a certain point the GALILEO and BeiDou bands are mixed up.

By the way, is SBAS supported by Mi 9?

The support should exist on all Snapdragon 855 devices.

I must say that Mi 9 is a very disappointing device to me in terms of GNSS.

The small typo that is in a library somewhere isn't that big of a problem honestly. I can get a GPS fix inside quite a few buildings and the accuracy outside is 2.5m easily 99.99% of the time when I've recorded long trips. I did however have an LG device before so my standards might be a bit different.

The unsupported accumulated delta range data is the worst thing...

I'm curious, what can you do with that data? Also have you tried the MIUI beta maybe it exists there?

What channels do you use to communicate with Xiaomi?

The built-in "Feedback" application in settings. If enough people report a problem then they fix it. They also have a forum, if someone posted about the GNSS issues there I'd jump on the train and also comment on it.

[PedroVenancio]

Thanks for your feedback @TaaviE !

The unsupported accumulated delta range data is the worst thing...

I'm curious, what can you do with that data? Also have you tried the MIUI beta maybe it exists there?

Accumulated Delta Range is necessary for post-processing workflows. Please, see this issue for more info: https://github.com/barbeau/gpstest/issues/268

I can get a GPS fix inside quite a few buildings and the accuracy outside is 2.5m easily 99.99% of the time when I've recorded long trips.

I also have this kind of accuracies, which is great for autonomous use. But with post-processing, we can get sub-metric accuracy.

[TaaviE]

Accumulated Delta Range is necessary for post-processing.

I found that information but I haven't found what (error) it corrects?

[PedroVenancio]

Accumulated Delta Range is necessary for post-processing.

I found that information but I haven't found what (error) it corrects?

https://developer.android.com/reference/android/location/GnssMeasurement.html#getAccumulatedDeltaRangeMeters()

https://insidegnss.com/generating-carrier-phase-measurements/

[benzfish]

Mi9 shows both L1 and L5 at the column L1, its a bug. Xiaomi will fix this bug in the next release.

[mimichris]

Hello, On the latest version MI9, MIUI 10.2.14 of this month, no change, I wonder if they will really change something ....

One question: Can these bugs affect GPS accuracy ?

[TaaviE]

They fixed some but not GALILEO, I submitted an error report, please do the same and ask others as well.

[Bolognesus]

Mi 9T pro, same issue.

[DominicHayes]

Impressed by the performance even at 10000m and 800kph, but the labels still need attention. 9T Pro.

[barbeau]

@DominicHayes Thanks for the screenshot! The carrier frequency of 1602 MHz for Galileo and Beidou is an error by Xiaomi - there are no real GNSS signals broadcast on these frequencies from these systems. Could you please report this bug to Xiaomi and see what they say?

Has anyone else gotten a response from Xiaomi on this recently?

[TaaviE]

@DominicHayes

I got 52 satellites on almost the same altitude with a Mi 9, also has a few weird Galileo frequencies.

I'm waiting for MIUI11 to see if Xiaomi has fixed it there.

[PedroVenancio]

I've just updated to MIUI 11 and it seems that this issue was fixed:

[TaaviE]

Did the same, seems fixed indeed.

[mimichris]

Hello, I do not have the possibility of this update yet, when did you have it? Thank you.

[PedroVenancio]

Hello, I do not have the possibility of this update yet, when did you have it? Thank you.

@mimichris It landed in my Mi9 last night.

[mimichris]

Hello,Maybe I had not looked at the possibility of updating, I just made this update and I no longer have the satellite data error. mimichris