About CTE sampling

RE: About CTE sampling

zhangpc — Mon, 29 Jan 2024 06:14:20 GMT

Hi gentleman,

That is a nice Q&A, and I am also trying to implement some non-standard AoA methods based on nRF52833. Mark this post!

RE: About CTE sampling

ethan-l — Tue, 16 Jan 2024 01:18:41 GMT

Thank you for your answer

RE: About CTE sampling

Simonr — Mon, 15 Jan 2024 09:13:37 GMT

Got an answer pretty quickly after my last reply:

1. The data is sampled prior to de-whitening.

2. The information needed to set this up can be read out rom the PS + the nWP036 Whitepaper. The key part would be to select which event to use to start the sampling procedure. This is user selectable from what I can tell.

Best regards,

Simon

RE: About CTE sampling

Simonr — Mon, 15 Jan 2024 08:33:35 GMT

I have forwarded these questions to the developers working on Direction Finding. Currently waiting for answers. Thank you for your patience!

Best regards,

Simon

RE: About CTE sampling

ethan-l — Fri, 12 Jan 2024 02:05:18 GMT

We assume that this method is done by sampling the binary byte stream of a particular field,So we're going to look at binary sequences of particular fields to see if there's any pattern. Now there are two questions:
1. The transmitted data has whitening and dewhitening processes, I want to know if the IQ sample was taken before dewhitening or after dewhitening, which results in a completely different data flow.
2. As you mentioned earlier, need to determine what to trig from and how much delay from the trigger point to the first sample to be taken. If DFEINEXTENSION is set to 0, It will sample from the packet payload,How do I set the offset so that the first sampling point is sampled in a particular field? I only found TSAMPLEOFFSET and TSWITCHOFFSET, By adjusting this, whether the payload offset several individual bytes is supported?

RE: About CTE sampling

Simonr — Thu, 11 Jan 2024 13:59:17 GMT

I can't say I'm confident enough to go into detail on the principle of how this proprietary IQ sampling works, and it's expected of customers that want to use this that they get the understanding of this on their end, as we don't really have any samples or drivers/libraries for it. Since it would be outside of the BLE specification, I don't think it would be correct to compare it to a specific BLE version, but I don't think it is restricted to the nRF52833/nRF52820 for example, since CTE won't be required.

Best regards,

Simon

RE: About CTE sampling

ethan-l — Thu, 11 Jan 2024 02:37:06 GMT

Hi,

Thank you very much for your answer. A proprietary mode is achieved by configuring Radio registers, as described in the Direction Finding White Paper? As you said, we understand that some manufacturers are already using this solution, such as quuppa, The IQ is obtained by adding a specific sampling field to the broadcast packet instead of sampling the CTE. I would like to ask how special fields in the broadcast can be set to sample valid IQ values, and what is the principle? For broadcast equipment, is it required to be BLE 4.2, which can be 52832/52810 instead of just 52833?

RE: About CTE sampling

Simonr — Wed, 10 Jan 2024 06:47:13 GMT

Setting the DFEINEXTENSION bit to 0 and doing I/Q sampling in the payload will not be supported by Bluetooth and is for proprietary modes only. How to use it is described in the PS (the output from the radio will be the same as for normal CTE, you just need to determine what to trig from and how much delay from the trigger point to the first sample to be taken), but as it will only be used for proprietary solutions it's difficult to describe in the likes of a white paper. There are people that have already implemented this by using the whitepaper for "normal" operations, and adhering to the PS for timing of the start, etc.

Best regards,

Simon