Nav Message Types & SBAS in GpsTest DB

ndo...@gmail.com

unread,

Jun 8, 2021, 4:42:08 PM6/8/21

to GPSTest

Thanks Sean for the wonderful GpsTest database, and your series of articles! It's an amazing resource!

A couple thoughts on possible DB improvements.

Recommend adding a column containing an ordered, enumerated list of GPS Navigation message types received (if NAV_MSGS supported). For instance, I'd like to know if a phone actually sends the L5 CNAV message, vs just the LNAV from L1.

https://developer.android.com/reference/android/location/GnssNavigationMessage

With respect to SBAS, if a phone supports one SBAS it should support them all. I totally understand why you're doing it the way you are, because of the various software/firmware lockouts, etc. I was thinking a column for specifics on unexpected behaviors w.r.t. SBAS usage. Like it utilizes EGNOS, but not WAAS or MSAS. This would indicate a software/firmware implemented lock-out, either by PRN (unlikely since PRN assignments change over time) or via the "Service Provider ID" in SBAS Message Type 17. https://www.gps.gov/technical/prn-codes/

It also might be interesting to understand if the phone is able to process one SBAS service message, and not the other. A MSG Type 27 (EGNOS), but not a 28 (WAAS)....or vice versa. EGNOS' plan as I understand it, is to migrate to utilizing a MSG Type 28 eventually, deprecating its usage of T27 completely.

For SBAS, it would be nice to know if the chipset utilizes SBAS satellites as additional ranging sources, or just as a message provider. I think you mentioned in one of your articles that may not be possible given the current API?

I really like that you included the SBAS carrier frequencies column. I know WAAS is currently broadcasting the same message set on both L1 & L5 on all three GEOs. However the L5 message stream will diverge from L1 in the future when WAAS upgrades to DF operations. Once SBAS's begin sending corrections for Galileo (currently only GPS), I wonder if the API will provide a mechanism for discerning that, or if you'll have to cross-reference against known PRN's.

Thanks again!

V. Kelly Bellis

unread,

Jun 9, 2021, 8:27:02 AM6/9/21

to GPSTest

Great ideas!

Maybe optionally (?) redact lat/long data - certainly would be cool to map where GPSTest is in use!

Be sure to include sentences that, whenever available, provide ellipsoidal height, ortho height, geoidal separation, datum information

$GNGGA

$GNGNS

$GNDTM

$PTNL,BPQ

$PTNL,GGK

$GNADV

Kind regards,

Kelly

ndo...@gmail.com

unread,

Jun 9, 2021, 9:27:54 AM6/9/21

to GPSTest

Quick correction: enumerated list of GNSS Navigation message types received.

Sean Barbeau

unread,

Jun 9, 2021, 2:16:53 PM6/9/21

to GPSTest

> Thanks Sean for the wonderful GpsTest database, and your series of articles! It's an amazing resource!

Glad you find it useful!

>Recommend adding a column containing an ordered, enumerated list of GPS Navigation message types received (if NAV_MSGS supported). For instance, I'd like to know if a phone actually sends the L5 CNAV message, vs just the LNAV from L1.

https://developer.android.com/reference/android/location/GnssNavigationMessage

Good idea, I opened an issue at https://github.com/barbeau/gpstest/issues/522.

Please follow along there. This is tricky to test as I don't currently have access to a device that actually supports navigation messages. If anyone does and you're willing to test please chime in on the above issue.

>With respect to SBAS,...

These are good ideas, but I'm reluctant to add more context surrounding SBAS until Android provides more information about how SBAS is handled by the platform.

Current Android support for SBAS in the APIs accessible to GPSTest is limited to flagging a satellite as "SBAS" (https://developer.android.com/reference/android/location/GnssStatus#CONSTELLATION_SBAS) vs one of the GNSS constellations. I actually have to do manual mapping based on the ID to the SBAS constellation myself:

https://github.com/barbeau/gpstest/blob/master/GPSTest/src/main/java/com/android/gpstest/util/SatelliteUtils.java#L130

Android doesn't specify whether an SBAS is "in use" for ranging or applying corrections, and I've also observed SBAS satellites appear and disappear unexpectedly in the list of available GnssStatus objects (satellites). Because Android can get almanac and ephemeris data from other assistance sources, it's currently not clear how these interact with SBAS. https://github.com/barbeau/gpstest/issues/269 for more details.

I'd suggest opening a feature request issue on Android for more specific APIs/data you'd like to see surrounding SBAS and hopefully Google will take a look:

https://issuetracker.google.com/issues/new

Sean

Sean Barbeau

unread,

Jun 9, 2021, 2:21:08 PM6/9/21

to GPSTest

> Maybe optionally (?) redact lat/long data - certainly would be cool to map where GPSTest is in use!

Currently you can see which country a submission originated from in Column AA "User location country code". For privacy reasons I really don't want to increase the precision greater than this.

> Be sure to include sentences that, whenever available, provide ellipsoidal height, ortho height, geoidal separation, datum information

Could you please open an issue at https://github.com/barbeau/gpstest/issues explaining the use case for each of these NMEA sentence types?

I'm not necessary against adding them, but I just want to be sure there is a clear reason and use case for adding more columns to avoid database bloat.

Sean

Sean Barbeau

unread,

Jul 1, 2021, 6:08:46 PM7/1/21

to GPSTest

ndoggac,

Do you have access to device that supports navigation messages? If so, could you please log some to a file and share?

That would help me implement this. Thanks in advance!

Sean

On Tuesday, June 8, 2021 at 4:42:08 PM UTC-4 ndo...@gmail.com wrote:

Reply all

Reply to author

Forward