CSI units

[sderonneTCS]

In which units is CSI calculated? I see in csi_func that you can retrieve imag and real per subcarrier, but how can I interpret these values?

[xieyaxiongfly]

That's a good question.

CSI doesn't have units.

The complex number of CSI is related to the Constellation map of Modulation. Generally, it tells the variation of a signal after going through the wireless channel.

[navidtadayon]

Yaxiong, I'm also wondering about the unit of CSI amplitude and couldn't understand your explanation. According to our experimentations, CSI from Atheros doesn't seem to be SNR as the range is about a few hundreds at best, neither it is snr in dbm...

Can you shed more light?

[xieyaxiongfly]

Definitely not SNR. I am so sure that CSI amplitude is not SNR for any Wi-Fi cards. You can look at the website I have updated first http://wands.sg/AtherosCSI/ and you may get a better understanding about the amplitude.

On Tue, Jun 12, 2018 at 12:07 PM, Navid [email protected] wrote:

Yaxiong, I'm also wondering about the unit of CSI amplitude and couldn't understand your explanation. According to our experimentations, CSI from Atheros doesn't seem to be SNR as the range is about a few hundreds at best, neither it is snr in dbm...

Can you shed more light?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/xieyaxiongfly/Atheros-CSI-Tool/issues/23#issuecomment-396459545, or mute the thread https://github.com/notifications/unsubscribe-auth/AF-RQ-ozFaQT-OcmX7UPewpCdfOkrJXuks5t7z6QgaJpZM4StsMF .

--

*-- Yaxiong Xie / *谢亚雄

School of Computer Engineering

Nanyang Technological University, Singapore

[navidtadayon]

Thanks for the link. It was useful. Still this doesn't answer the question. Even though CSI amplitude at each subcarrier is not SNR, it has power in it. The reason I'm claiming this is better explained in the following: As you noted in your link, CSI amplitude (vs subcarrier) has a time-changing offset, which most likely is caused by the variable AGC gain. since each rx chain uses an independent AGC component, signals on each chain is amplified independently. Now if you form the metrix \zeta=CSI_{2,1}/CSI{2,2} where the receive index is fixed, you see that all that fluctuations go away. That indicates two points: (1) signal strength (or power) is in CSI amplitude (2) it's not in logarithmic scale, otherwise we wouldn't be able to remove it by "division".

Now the fact that what other factors contribute in CSI amplitude is really mysterious to me. We are doing TOA ranging with your tool at UofT (and before that I did the same at MIT) and we were successful to do it with phase. Still can't wrap my head around the unit of CSI amp...

[xieyaxiongfly]

What do you mean by \zeta=CSI_{2,1}/CSI{2,2}? Can you explain with more details?

[ghost]

Navid, Yaxiong,

I understand both of you are probably moved on to other exciting projects by now, but I am wondering if you might be able to quickly address some further questions on the CSI amplitude values, or at least point one in the right direction. In my current project, we are hoping to use the CSI amplitude values in combination with the phase values (of which we thankfully have a stronger understanding) for our calculations. Did either of you discover further understanding of how to compensate for the AGC gain while scaling the amplitudes? Can the RSSI values be used as a corrective multiplier in this scaling process?

Thank you both, we appreciate all of your contributions to this forum to date.

[navidtadayon]

I can just briefly weigh in on that: I remember that amplitude of CSI is uncompensated meaning it carries the gain of AGC. RSSI is also not a perfect measure for compensating it. Best if you use phase information unless you exactly know the AGC gain.

---

Bests Navid

On Mon., Aug. 26, 2019, 11:46 a.m. shaneearley10 [email protected] wrote:

Navid, Yaxiong,

I understand both of you are probably moved on to other exciting projects by now, but I am wondering if you might be able to quickly address some further questions on the CSI amplitude values, or at least point one in the right direction. In my current project, we are hoping to use the CSI amplitude values in combination with the phase values (of which we thankfully have a stronger understanding) for our calculations. Did either of you discover further understanding of how to compensate for the AGC gain while scaling the amplitudes? Can the RSSI values be used as a corrective multiplier in this scaling process?

Thank you both, we appreciate all of your contributions to this forum to date.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/xieyaxiongfly/Atheros-CSI-Tool/issues/23?email_source=notifications&email_token=AGR2KJDBXGMSQQAVDZVXOPDQGP3DXA5CNFSM4EVWYMC2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5EYWOI#issuecomment-524913465, or mute the thread https://github.com/notifications/unsubscribe-auth/AGR2KJBOFH2X5RV57ENMIWTQGP3DXANCNFSM4EVWYMCQ .

[ghost]

Thank you for weighing in with your thoughts Navid -- this is very helpful.