bluetooth 5 interference cuts audio/instructions out

[tt8484]

Problem: while listening to music over bluetooth A2DP profile from gps (zumo 660, bluetooth 2) to a headset (sena 20s, bluetooth 4.1), the music is interrupted when crossing path with other people on the road.

Additionally, when music is not playing (assuming the failure occurred silently), the gps instructions are not sent either. Powering the headset down and back up causes the gps to rapidly fire all instruction that were queued since the failure.

This started in fall 2017 but got worse in 2018 and in 2019 lately it is almost certain to happen (maybe 3 times out of 5 encounters).

Hypothesis: since 2017, bluetooth 5 (BT5 from here on) has been introduced in smartphones. It's adoption increased with more and more people buying newer smartphones. I suspect these modern BT5 phones are emitting BT packets that cause the headset or gps to loose connection (or something causing the same symptoms).

What could be happening here? I suspect a few things:

a) BT5 has an extended advertising (255 bytes instead of 37). Obviously the advertising is connection-less (broadcast). Could this crash a BT2 stack not expecting more than 37 bytes?

b) Could the extra advertising bytes be mistaken for some RFCOMM control flow? This seems not very likely because the unique id would have low chance of matching, not the 3/5 I experienced.

c) could the BT5 simply pollutes the airwaves enough to mask a "clear-to-send" (CTS) control flow from the headset? This seems unlikely given the spread spectrum frequency hopping technology.


-have you experienced this?

-what else?

-Do you know of any other keywords or articles describing this problem?

-how should a device defend itself (or facilitate recovery for the other) to avoid the problem? I can imagine a timer which resends the CTS or reestablish connection within 1 second... I fear we won't win the battle if we ask apple/google/samsung to stop using BT5...!

-how do we get garmin or sena to fix the issue? (besides the bug reports I already sent)? If you have such issue, I need you to file your own bug report to them, to add pressure.

[alan sh]

I doubt that Garmin will fix this on their 660 - that's way out of date now.

Can you try with a modern Garmin device or a dfferent headset? See if you can isolate the issue?

Alan

[tt8484]

I doubt that Garmin will fix this on their 660 - that's way out of date now.
Can you try with a modern Garmin device or a dfferent headset? See if you can isolate the issue?
Alan

Garmin did fix a major date bug just last month on the 660, so they can do it again for something worse, especially if they get some help to get started. I think positive, they never let me down.

Unfortunately I don't have any other gps or headset to test.

[tt8484]

I managed to revive an old headset with a new battery and the problem was not showing.
So this is a problem with the sena 20s headset.