Scanner performance test feedback.

[cl]

Hi.

I'm currently trying to scan beacons with timeslot example.

However, I found that there're more than 20% of loss while scanning beacons broadcast.

I think that's huge even if broadcast has no guarantee to be transferred.

I set up 1 beacon to broadcast with specified UUID in a short interval.

The interval is 20ms and I achieved this with repeating start & stop manually.

The beacon broadcast for 256 time and each time change minor value of packet. (0x00 to 0xFF)

Then, I set a scanner to scan beacon broadcast packets with specified UUID.

(I make it compare UUID of each packet manually.)

For the physical environment, I placed a beacon and a scanner close to each other. (less than 20cm)

And then, I checked out how much loss happened in the scanner side.

For various combinations of scan interval & window value, loss rate always exceeds 20%.

I wonder if this is expected normal performance.

(This question is related to previous question )

Can I get a code review for this test?

Or can I get a verified result & sample code for testing?

Thanks in advance.

[Ole Bauck]

I cannot see any code, but I would guess it is correct. Unfortunately packet loss will happen. In your case it looks like it is because interference. Interference could be from your mobile, wifi router, wireless mouse etc, as many wireless items use the 2.4 GHz band.

You can go to a place where you know there are less interference and place the beacon and scanner close to each other. See if the result is different.

You can try with multiple scanners and see if they miss the same advertisement packets.

You can use DTM to test the channels, see application note and DTM example in SDK.