CS Configuration - Issues with Movement

[witc]

Hi,
when I am trying to set up channel sounding, I am looking for the best parameters for CS configuration. The main problem I have is that even with slow movement (1 km/h) and at a maximum distance of 6 meters, many procedures do not finish successfully. There are many wrong data points and the result cannot be calculated. I am testing this in a room.
Does the CS procedure need a higher RSSI level to keep reliable measurements, or are the RSSI values the same as with normal BLE communication enough?

If I use what is in the example cs_ras_initiator, I see a big error rate - the results are not available at all.
If I increase min_procedure_interval to hundreds (for example 3*160), it starts to work quite well - it works reliably, but only until I start moving the initiator a little. Then, I think there is some data loss, and the whole procedures have invalid data.
I do not know if there is a way to make it more robust against movement. Maybe it is also because of a bit worse hardware - I have my own custom boards. I use 1 TX and 1 RX antenna.

Thank you, Jan.

[Amanda Hsieh]

Hi, 

CS relies on new features of the radio (PHY), features of the controller, security measures, and measurement procedures needed to collect raw data. The conversion of raw data into the resulting distance using dedicated algorithms is performed at the application level and is outside the scope of the Bluetooth specification. 

The default sample was designed to validate the data flow only, and it's not robust against movement or multipath. Parameter tuning can help, but may not fully resolve issues with movement or reflections. See my colleague's suggestion in this post. Hardware quality and antenna design are important factors. If possible, test with Nordic DKs to rule out hardware issues. 

Regards,
Amanda H.