Channel sounding sample distance accuracy

RE: Channel sounding sample distance accuracy

witc — Thu, 11 Sep 2025 08:28:22 GMT

Hi, fromy my experience (I am also elaborating with CS) is problem with multipath of RF signal. Are you indoor? Also I am not sure if making calibration for ifft is needed..

I am trying more Anetanna path, but it does not work very well for me..

RE: Channel sounding sample distance accuracy

Tom.Fu — Thu, 11 Sep 2025 01:55:13 GMT

Hi SwRa

After I increased the TX power to 7dBm of maximum for both initiator and reflector(both are nRF54L15 DKs), the initiator detect distance can exceed 18 meters, but the distance measurement accuracy has declined when the distance between initiator and reflector exceeds 25 meters.

For example:

The ifft is about 23.9 meters when the distance between initiator and reflector is 30 meters.

The ifft is also about 23.9 meters when the distance between initiator and reflector is 35 meters.

Is there any approach to optimize it? Many thanks.

RE: Channel sounding sample distance accuracy

SwRa — Wed, 10 Sep 2025 11:33:46 GMT

Hi Tom,

[quote user="Tom.Fu"] Can the channel sounding samples be configured to increase the Tx power?[/quote]

Yes, the channel sounding samples can be configured to increase the TX power. The SoftDevice Controller API provides the function sdc_default_tx_power_set(int8_t requested_power_level) to set the default radio TX power level. You can also do this by setting CONFIG_BT_CTLR_TX_PWR_PLUS_X option in your prj.conf

For example, to set the TX power to 8 dBm, you would use sdc_default_tx_power_set(8); or set CONFIG_BT_CTLR_TX_PWR_PLUS_8=y in you prj.conf .

Best Regards,

Swathy

RE: Channel sounding sample distance accuracy

Tom.Fu — Tue, 09 Sep 2025 09:45:17 GMT

Hi SwRa

Thank you. The distance accuracy gets much better under your suggestions. But the initiator can’t detect the reflector when the distance between them exceeds 18 meters. Can the channel sounding samples be configured to increase the Tx power? Many thanks.

RE: Channel sounding sample distance accuracy

SwRa — Thu, 04 Sep 2025 12:55:15 GMT

Hi Tom,

Yes, the IFFT value is the measured distance.

[quote user=""] ifft distance error is relatively large.[/quote]

The current channel sounding samples in NCS are not fully optimized for accuracy. The phase slope algorithm, for example, is not recommended for performance evaluation. This is discussed in the following ticket: (Missing Forum Thread)

Make sure you are using the latest NCS version (v3.0.1 or newer), as there have been some improvements in the algorithms, though fundamental limitations remain.

Regards,

Swathy