nrf54L15,channel sounding test

[wjw2025]

Hi

The nRF54L15-DK suite tests the accuracy of channel sounding, all ncs versions are v3.1.1.  I am using the channel_sounding_ras_initiator and channel_sounding_ras_reflector examples under the ncs\v3.1.1\nrf\samples\bluetooth path for verification.  

Power-on serial port log ：

Launch nRF Connect SDK v3.1.1-e2a97fe2578a ***  

Use Zephyr OS v4.1.99-ff8f0c579eeb ***  

1、There are many such mistakes：

[15:35:50.484]收←◆I: Ranging data dropped as peer ranging counter doesn't match local ranging data counter. (peer: 26, local: 24)

[15:35:51.116]收←◆I: Dropped subevent results. Waiting for ranging data from peer.

2、What exactly does the ifft, phase_slope in the following log represent and how can it be converted into distance？

 Distance estimates on antenna path 0: ifft: 2.840086, phase_slope: 4.802810, rtt: 11.392113

3、Does the detection accuracy test of this channel have a significant impact on the placement of the antenna on the board?

4、How are the test results of this version from your official side?  

Thank you! 

[tesc]

Hi,

What exactly does the ifft, phase_slope in the following log represent

The numbers are distance estimations, based off of three different calculation methods: Inverse fourier transform (ifft), derivative of phase with respect to frequency (phase_slope) and round-trip time (rtt). Read more about the methods at Bluetooth Channel Sounding Distance Estimation. I see that from your numbers, the estimates are widely different between the methods. At what actual distance was this estimate performed?

Does the detection accuracy test of this channel have a significant impact on the placement of the antenna on the board?

I am sorry, but I do not quite understand what you mean. Can you try to rephrase the question, or elaborate?

How are the test results of this version from your official side?

I am looking into what is our results and recommendations, and expect to get back to you with that after the week-end.

Regards,
Terje

[tesc]

Hi,

1、There are many such mistakes：

[15:35:50.484]收←◆I: Ranging data dropped as peer ranging counter doesn't match local ranging data counter. (peer: 26, local: 24)

[15:35:51.116]收←◆I: Dropped subevent results. Waiting for ranging data from peer.

Those indicate that ranging data of the given index could not be transferred to the initiator before new ranging data arrived from the CS procedures, so it had to be dropped. The reason could be that the CS procedure interval is too low in relation to how much data each CS procedure produces, or that the connection is dropping packets. It might help to increase the number of ranging data buffers in the ranging responder, see CONFIG_BT_RAS_RRSP_RD_BUFFERS_PER_CONN.

tesc said:

I am looking into what is our results and recommendations

Theoretical accuracy of Channel Sounding, according to Bluetooth SIG, is typically +/- 50 cm. See e.g. Bluetooth Channel Sounding: How it works and what it means. Our solution in nRF Connect SDK is of experimental status and has not been tested for accuracy, but we do have some suggestions based on experience.

What we have seen, is pretty consistent measurements if calibrated. To calibrate, estimate distance at a short distance in order to find the offset from the real distance, and then apply that offset for correcting other measurements. For phase based ranging, it should be perfectly possible to achieve 0.5 meters accuracy up to 10 meters, and 1 m accuracy up to 20 meters. RTT can be used for longer distances and to confirm the IFFT results, but is not accurate on its own.

Any questions regarding when our distance estimation libraries will have production status revolves around roadmaps and future plans, for which you should reach out to sales. Please use the form under "Sales related questions" on https://www.nordicsemi.com/About-us/Contact-Us, or contact the responsible for your area directly if you know who they are and have their contact details.

Regards,
Terje