www.st.com

[Francesco Virlinzi]

Hi all

Have a look @

https://www.st.com/resource/en/application_presentation/teseo-3-quick-testing-guide.pdf

it's an easy way to test the connection to the Teseo III server based Assisted GNSS without special equipment and/or tool.

[EHuse]

Okay, now I'm a little overwhelmed. I think I have found the X-CUBE-GNSS1 package and downloaded it. But there's a lot of code there.

Are you saying that to use A-GPS we need to link in a pre-built library into our code base that will parse the JSON responses?

Our board uses a QCA9558 processor and we are running a version of OpenWRT on it.

[Francesco Virlinzi]

Ciao

You should follow the A_GetPos project.

> Are you saying that to use A-GPS we need to link in a pre-built library into our code base that will parse the JSON responses?

Yes. Correct.

> Our board uses a QCA9558 processor and we are running a version of OpenWRT on it.

Sorry I don't know this processor neither OpenWRT.

Is QCA9558 Arm-Cortex based?

Thanks

Ciao

Francesco

[EHuse]

The QCA9558 is a MIPS 74Kc processor, not an Arm. So I don't think we can use the prebuilt library, which is unfortunate.

Are there any other options?

[Francesco Virlinzi]

Ciao

In this case NDA is the only option we have.

Regards

Francesco

[EHuse]

Hey again,

So I will talk to my manager about getting an NDA signed. Do you have a contact person he should speak with?

Thanks much,

Eric

[Francesco Virlinzi]

Ciao

You can contact: michael.slade@st.com

He is already informed.

Ciao

Francesco

[EHuse]

Hello again,

So I now have access to the necessary code and am currently working with it to figure out how to decode things. I have a few more questions. 🙂

1. Is the extended ephemeris data (the 2 week data) also referred to as the Almanac data? Is this Almanac/extended ephemeris data the same data that is cleared if I send a $PSTMCOLD,0xF command to the chip?
2. For the Real Time Ephemeris data request, does that require knowing some sort of rough position to begin with? (Like from a wifi or cellular id?). The quick testing guide has some examples of RT Assistance requests but they don't seem to have any specification of initial location.

Thanks much for any information.

[Francesco Virlinzi]

Ciao

On (1):

The autonomous Assisted GNSS has an internal database of ephemeris data (we call it "synthetic" ephemeris) and the predictive Assisted GNSS is able to extend the 'valid' time (from 4 day to 2 weeks) of autonomous assisted GNSS.

When you perform '$PSTMCOLD,0xF' you are clearing the real almanac and the real ephemeris data... not the "synthetic' ephemeris which remains as they were.

If you want clear the "synthetic" ephemeris you have to use the command $PSTMSTAGPSINVALIDATE which clears the "synthetic" ephemeris

On (2):

>  does that require knowing some sort of rough position to begin with?

On it doesn't because in case of Real-Time ephemeris you are downloading all ephemeris of constellation (not just a subset...) for this reason wifi or cellular-id are not required.

Take care the the Real-Time downloaded data will be used as 'real ephemeris and real almanac' (because they are really real 😉 this means they are cleared with a ' $PSTMCOLD,0xF'

Hope this helps

Regards

Francesco

[EHuse]

Hey again,

So I'm trying to experiment with taking the printed JSON response and feeding that into my ported code version. Right now it is decoding to all 0s. 🙂 With the example output in the testing guide, are all the sequences of "\/" actually using a backslash to escape the forward slash in the string? So the actual JSON data is just one forward slash? Or is it actually "\/" in the data? I ask because I never see just a single forward slash in the code anywhere.

Also, so the RTAssistance JSON request is parsed out to generate $PSTEPHEM data. And this is the actual real time ephemeris data.

Then the predictive JSON request is the one that used the seed and block commands to setup the "synthetic" or predictive Assisted GNSS?

I do appreciate all the help, I'm making progress but just don't have all the code figured out yet.

Eric

[Francesco Virlinzi]

Ciao

> With the example output in the testing guide, are all the sequences of "\/" actually using a backslash to escape the forward slash in the string?

Yes. They are.... mandatory to translate the J-SON request-sentence in standard C-Code (inline with the C-Code escape rule).

Therefore (for example) the json request-data for predictive:

[{ "ee": {"version": 8, "constellations": ["gps", "beidou","galileo"], "seedAge": 0} } ]

becomes in C-Code:

"[ { \"ee\": {\"version\": 8, \"constellations\": [\"gps\", \"beidou\", \"galileo\" ], \"seedAge\": 0} } ]"

which seems something crazy... but working 😉

> Then the predictive JSON request is the one that used the seed and block commands to setup the "synthetic" or predictive Assisted GNSS?

Yes. Correct.

Seed (dowloaded from served) and block commands are required to fill the synthetic ephemeris for Predictive GNSS.

> Also, so the RTAssistance JSON request is parsed out to generate $PSTEPHEM data. And this is the actual real time ephemeris data.

Yes. Correct.

RT-assisted data FILL the real-ephemeris (RT data are NOT a prediction... They are REAL satellite ephemeris data downloaded in one-shot for the whole constellation)....

Ciao

Francesco