Invalid values on RawSignalStrengthInDBm on BLE advertisment

A.P 1

Hi!

We're developing a UWP app, which functions as our BLE central, and 'speaks' to many BLE peripheral units. When a connection is formed, the RSSI value for that central-peripheral connection is logged and reported.
We're seeing two flavors of the same problem, The problem is that some portion of the RSSI values we're reading make no sense.
first flavor- RSSI values we get are positive. Examples: 79, 80 etc.
second flavor- RSSI values we get are unreasonably low. Example: -127. This is well below the link budget and doesn't seem like a 'real' RSSI value.

These values are received from Windows.Devices.Bluetooth.Advertisement -> BluetoothLEAdvertisementReceivedEventArgs when receiving events from BluetoothLEAdvertisementWatcher. So they seem to be originating straight from the driver. No manipulation on our side.

Both of these problems are not constant. We'll get reasonable values for a while (-40, -48 etc) and then an anomaly will be recorded, and back to 'normal' values.
We suspect the fault lies with the windows driver, as when we try to switch dongles (different manufacturers with different chipsets), we still see the same strange values pop up every once in a while.
We use Windows.Devices.Bluetooth under UWP (universal windows platform) on Intel NUCs.
Updating windows to the latest version, and updating the BL drivers etc did not solve the problem.

Has anyone experienced these kind of issues before?
Thanks!

Fay Wang - MSFT 5,196 Reputation points

2020-08-18T06:50:51.197+00:00

When you try to receive the RSSI values again, would you follow this guidance to provide your Bluetooth log for us to check? In addition, here is an official sample about BLE advertisement, if you test it, will the same issue occur? For more details about RawSignalStrengthInDBm, you can refer to this document.
A.P 1 Reputation point

2020-08-19T13:14:47.057+00:00

Hi FayWang!

I'm attaching alink to GDrive, since I can not upload the log files as they are too large.

https://drive.google.com/file/d/1U_cjTfx62gXsuLunPV3jPF50v1PlDFCH/view?usp=sharing

You'll find two files- one with a few 'positive value' RSSI reads. The second with a -127 RSSI read which should be towards the end of the run.

Please let me know if you need anything else from me.

Thanks!
A.P 1 Reputation point

2020-08-24T07:51:45.54+00:00

Hi FayWang!

Just to let you know, We've updated our bios, BlueTooth and other drivers to the latest release from Intel for all our NUCs (I'm referring to the drivers released on August 2020).
We see some improvement in the frequency of 'bad' RSSI values following these updates, but the issue is not completely resolved. If before we saw these values once every 20 readouts, we now see them about one every 50 read outs.

I'm hoping you can point us in the right direction using the log files I've linked.

Thanks!
A.P 1 Reputation point

2020-08-25T16:26:58.367+00:00

Hi!

Here is a link to the files relevant to 'Driver issues' as described in the link you've sent,
https://drive.google.com/file/d/1AgL8S3PJXTlk0xHcropQNRe7JpI_YF-0/view?usp=sharing
My apologies, I didn't spot that part before.
I've been able to catch the '-127' value of the RSSI read, you can find it in the files I've added. The positive RSSI value still eludes me

Please let me know if I missed anything else.
A.P 1 Reputation point

2020-08-26T08:41:20.877+00:00

Caught the 'Positive RSSI' issue as well.
See files here
https://drive.google.com/file/d/13HATuDgXWKru6KoQQePJccZAAEwbYgTR/view?usp=sharing
A.P 1 Reputation point

2020-08-26T15:20:41.083+00:00

I've been able to get an RSSI value of -127 using the official sample watcher. I've not observed positive values thus far.
A.P 1 Reputation point

2020-08-26T15:25:28.633+00:00

Caught a positive RSSI readout
Fay Wang - MSFT 5,196 Reputation points

2020-09-02T06:31:48.107+00:00

There is nothing "invalid" about negative values for the signal strength. It is quite possible that environmental issues can affect the signal strength, even for a short time, like other wireless devices causing interference for instance. You could see this document for more details.
Fay Wang - MSFT 5,196 Reputation points

2020-09-09T02:29:03.557+00:00

In addition, because of environmental issues, any single signal strength value should not be trusted, but instead it is better to look at an average of values over a period of time to get a better signal strength number to report.
A.P 1 Reputation point

2021-05-09T16:30:11.72+00:00

Hi FayWang!

Negative values are 'valid', yes. However they are valid when they are within a certain range. Is -127 within that range? that is below the sensitivity point of my hardware.

If I understand correctly, and I grant you right off the bat I am no expert, a -127 RSSI value is similar to getting a -280 Celsius reading from a thermometer. Yes, negative temperatures are possible, but not that low.