Question Regarding BLE Channel Sounding Ranging Failure Beyond ~1 Meter

[Mike Lyu]

Dear Sir or Madam,

We are currently evaluating and testing the **Channel Sounding feature on Nordic BLE devices**. During our measurements, we observed that the Channel Sounding results degrade significantly once the distance exceeds a certain range, which eventually prevents distance estimation. We would like to ask whether this behavior is expected, or if there are recommended configuration adjustments to improve the performance.

---

Test Environment

* Device: nRF54L15
* Environment: Open office space (LOS, non-anechoic)
* Device setup: Both devices are static, with fixed antenna orientation
* Channel Sounding configuration: Enabled Channel Sounding with 72-channel frequency sweep
* Sample used: `samples/bluetooth/channel_sounding_ras_initiator`
* Configuration status: No modifications made; all default sample settings are used

---

Observed Behavior

* When the distance is "below approximately 1 meter", Channel Sounding operates normally and ranging results can be obtained.
* When the distance is "greater than approximately 1 meter", signal quality frequently degrades, and the channel sweep results are often incomplete.
* Due to insufficient valid Channel Sounding data, the ranging algorithm is not executed, resulting in no distance output.
* We also observed that when the distance exceeds ~1 meter, the `tone_quality` field frequently reports `1 (fail)`, indicating that multiple channel tones do not pass the quality check.

---

Preliminary Assessment and Questions

Based on our observations, we believe this issue is likely not caused by the ranging algorithm itself, but rather by insufficient received tone quality at longer distances, leading to an insufficient number of valid channels.

We would like to ask the following questions:

1. Under the test conditions described above, is the significant degradation of Channel Sounding performance beyond ~1 meter an expected behavior or known limitation?
2. In general, which factors most commonly cause `tone_quality = fail` (e.g., SNR thresholds, CS procedure duration, TX power, timing configuration)?
3. Are there any recommended Channel Sounding configuration adjustments (such as TX power, subevent/procedure interval, main mode repetition, or quality thresholds) to improve measurement robustness and extend the effective ranging distance?
4. Is there an internal requirement for a minimum number of valid channels or quality criteria, below which the Channel Sounding result is considered invalid and the ranging calculation is skipped?

---

We would appreciate your guidance and experience to help us determine whether this behavior is an inherent system limitation or if there are further optimization opportunities.

Thank you very much for your support.

Best regards,
Mike

[Asbjørn]

Hello Mike,

No, the range isn't limited to within a meter. My initial question will be to ask what sort of HW you are using. 

Q1: Have you developed your own custom board or are the range testing done with the nRF54L15 DK?

Based on your observations I suspect that you are using a custom board. Would you be able to share with us the details of the boards as I suspect that the performance you are getting is a result of the HW.

Do you have two nRF54L15 DK available? What performance do you see when using those to for a comparisson test?

Best regards

Asbjørn

[Mike Lyu]

Hello Asbjørn,

Thank you for your clarification and suggestions.

To confirm, our current measurements are performed using nRF54L15 DKs on both the Initiator and Reflector sides, not a custom board. The observations previously described are based on DK-to-DK testing with default sample configuration and on-board antennas.

nRF54L15 DK ↓

Configuration ↓

   

Regarding your suggestion of using a different development board as the Initiator for comparison, we will proceed with this test setup and evaluate the ranging performance accordingly. Once we have the results, we will share our observations with you for further discussion.

If there are any specific test conditions, configurations, or metrics you would recommend focusing on during this comparison (e.g., PHY selection, TX power, antenna path, or logging details), please let us know.

Thank you again for your support.

Best regards,
Mike

[Mike Lyu]

Hello Asbjørn,

As a further update, we would like to share additional test observations.

We currently have four nRF54L15 DKs available and have conducted multiple ranging tests using different DK combinations (Initiator/Reflector pairings). Based on these repeated measurements, we observed that the maximum achievable ranging distance varies between individual DKs.

However, across all tested combinations, the Channel Sounding ranging functionality generally degrades and becomes unreliable beyond approximately 2 meters, after which distance estimation consistently fails due to insufficient valid Channel Sounding tone data.

This behavior appears to be reproducible across multiple test runs and DK pairings.

We would appreciate your insight on whether this level of unit-to-unit variation and the observed distance limitation are expected characteristics of the DK hardware or the default Channel Sounding configuration.

Best regards,
Mike

[Asbjørn]

Hello Mike,

I'm surprised that you are seeing this with the DKs you have. Can't really explain why and it makes me a bit concerned. May I ask which version of NCS and toolchain you have used to build the example? Sample used: `samples/bluetooth/channel_sounding_ras_initiator` I'd like to try and recreate the build here and see what it does. 

Best regards

Asbjørn