Talk:WiFi Sensing

new edits (for discussion)

I would suggest adding the following which points to the two very popular open source WiFi CSI tools:

There were opensource efforts from research community that led to published tools for extracting CSI measurement from commodity 802.11n NICs, including the Linux 802.11n CSI Tool from University of Washington and the Atheros CSI tool from Nanyang Technological University. --Macmobile (talk) 14:38, 28 September 2021 (UTC)

Wikipedia isn't a link directory, this article shouldn't contain a list of links to tools. - MrOllie (talk) 17:50, 30 September 2021 (UTC)

It's not only about listing links - the two tools are the most popular CSI tools used in academia, and the CSI is the essential basis for wifi sensing - which benefit whoever might be interested to explore. As a page for wifi sensing, I cannot think of a reason why the popular tools for retrieving CSI should not be listed. - Macmobile (talk) 08:40, 1 October 2021 (UTC)

We don't list things because they're popular, and Wikipedia is not a link directory - the purpose of Wikipedia is to host informative articles, not to provide links to tools users might want to 'explore'. - MrOllie (talk) 11:46, 4 October 2021 (UTC)

I think the major criteria that applies in wiki is "no original research", and we should keep the content as long as its provided information is verifiable and relevant (in the circumstance of wifi sensing, how to extract CSI is essential). Applying your above standard, much of existing content of this page may be subject to removal, e.g., the followings:

Much of the early academic research on WiFi Sensing was based on large Software-Defined-Radio (SDR) hardware, such as the Ettus Research USRP. SDR provided flexibility to perform custom operations which were previously impossible due to the close natured implementations of off-the-shelf Wi-Fi hardware.

One of the earliest announcements related to the commercialization of WiFi Sensing, a tool for extracting CSI measurements, was published in 2011. This tool was broadly available in the Wi-Fi industry and was compatible with an 802.11n network interface controller (NIC) from Intel.

In 2015, the first fully-integrated SDR on a single chip, known as the R10 (Radio10), was introduced by Cognitive Systems Corps. Its initial purpose was spectrum monitoring for cellular, Wi-Fi, and other Land-Mobile-Radio (LMR) services using a radio-frequency (RF) camera system to observe RF signals and their parameters from a pre-defined field of view. The chip had five custom CPU cores, four wireless receivers, and highly configurable dual multi-vector processors, giving the R10 chip significant capabilities in detecting and processing wireless signals in real-time. Once in production, Cognitive Systems Corps. focused on using the R10 to monitor the most prevalent spectrum, Wi-Fi signals, for motion detection. To further develop the signal processing algorithms, a major subset of the Wi-Fi MAC/PHY stack was implemented on the R10.

-Macmobile (talk) 06:37, 5 October 2021 (UTC)

Wikipedia is not an indiscriminate collection of information, we do not keep everything we can possibly source. Also, you have clearly seen that I disagree with this inclusion, there is no consensus for it to be in the article. You should not continue to restore your preferred version over the objections of others. - MrOllie (talk) 11:32, 9 October 2021 (UTC)

With my expertise knowledge in the area of wireless sensing and communication I believe my added content is very relevant to wifi sensing and of value. It is presented with neutral point of view, does involve no original research, and is verifiable, so I think it should be kept. - Macmobile (talk) 08:39, 13 October 2021 (UTC)

So you have said, and I disagree - Wikipedia is not a place to build a directory of tools. I have listed this for a third opinion, we'll see what they say. MrOllie (talk) 12:08, 13 October 2021 (UTC)

I came here looking for tools, and am kind of annoyed that I had to go to the Talk page to find them 69.201.39.112 (talk) 14:17, 22 December 2023 (UTC)

• Modify substantially. I gave it a few reads, and it seems like the bit about open source published tools adds something, but certainly not with external links, and I don't think mentions of the tools themselves add anything. It should also be copyedited for readability. Pyrrho the Skeptic (talk) 13:47, 13 October 2021 (UTC)

I suggest merge with the first paragraph in Commercialization (talking about the UWashington tool, which is a open source effort in academia and clearly not commercialization behavior) and move to under Academic section. Let me know whether the following edits make sense:

To REMOVE the first paragraph under "Commercialization";

To ADD the following under "Academic":

There were opensource efforts from research community that led to published tools for extracting CSI measurement from commodity 802.11n NICs, including the "Linux 802.11n CSI Tool" from University of Washington which is able to extract CSI from Intel 5300 NIC in a format that reports the channel matrices for 30 subcarrier groups, i.e., about one group for every 2 subcarriers at 20 MHz measurement or one in 4 at 40 MHz measurement, and the "Atheros CSI tool" from Nanyang Technological University which extracts CSI from 56 subcarriers for 20MHz measurement and 114 subcarriers for 40MHz measurement. Both tools have been popularly adopted for research and development in academia and industry.

Wait for a couple of days for comments and will implement on the main page - Macmobile (talk) 06:02, 14 October 2021 (UTC)

I don't know how you got from 'I don't think mentions of the tools themselves add anything.' to another proposal that mentions the specific tools. I continue to object to that wording. - MrOllie (talk) 11:37, 14 October 2021 (UTC)

I've put an attempted compromise version in the article, that mentions that tools exist without naming specific ones. - MrOllie (talk) 11:40, 14 October 2021 (UTC)

Seems like a good compromise. Contains the essential info. Pyrrho the Skeptic (talk) 15:41, 14 October 2021 (UTC)

I would say we should move this to under Academia - the tool and open source efforts were not for commercialization/profit purposes; second, I think we should also include the other CSI tool which got popular later and worked with Qualcomm NICs. Suggested edits (and move it to under Academia): -Macmobile (talk) 06:54, 16 October 2021 (UTC)

There were opensource efforts from research community that led to published tools for extracting CSI measurement from commodity 802.11n NICs from Intel or Qualcomm Atheros series Both tools have been popularly adopted for research and development in academia and industry. The primary motivation and uses were for wireless channel estimation, network troubleshooting, developing Multi-User MIMO systems, and understanding channel fading in Wi-Fi networks. 

The text I just added is under the 'academic' section already. We don't need this 'both tools' language, the second citation is redundant, and I see no reason to mention particular NIC vendors. - MrOllie (talk) 11:57, 16 October 2021 (UTC)

I see the paragraph about CSI tool is still under commercialization. There are two tools available and they work for different NIC models - I see no reason to mention one but omit the other. In my earlier edits I put a few words on the rationale of the CSI extraction - to collect CSIs (frequency responses) from different number of subcarriers for the different bandwidth - I think that will help readers understand the essentials of CSI measurement. -Macmobile (talk) 05:19, 18 October 2021 (UTC)

Fair enough, I removed the now redundant mention in the commercialization section. MrOllie (talk) 12:07, 18 October 2021 (UTC)

See my edits on yours - "Later efforts from the research community led to tools for extracting Channel State Information (CSI) measurements from commodity 802.11n NICs including Intel 5300 and Qualcomm Atheros series. The CSI measures the channel matrices from a variety of OFDM subcarriers that can describe the frequency responses of the 20MHz/40Mhz wireless channels used in 802.11n. The primary motivation and uses were for wireless channel estimation, network troubleshooting, developing Multi-User MIMO systems, and understanding channel fading in Wi-Fi networks." -Macmobile (talk) 07:56, 21 October 2021 (UTC)

Yes I saw, and as I said when you proposed it on this talk page I continue to object to that wording. MrOllie (talk) 12:07, 21 October 2021 (UTC)

There remains no reason for listing tools, or for the additional technical language. MrOllie's most recent restoration reflects the earlier compromise. Pyrrho the Skeptic (talk) 16:14, 21 October 2021 (UTC)

So how do we proceed? MrOllie is apparently happy with listing one tool but constantly object referring to the other tool. And my wording to explain how CSI works in the context of wifi sensing is objected - for no reason -Macmobile (talk) 02:33, 25 October 2021 (UTC)

We proceed with the sentence as it is. MrOllie's version did not contain a tool name, as far as I can see, and your technical explanation is esoteric and not useful to the general reader, in my opinion. Pyrrho the Skeptic (talk) 02:40, 25 October 2021 (UTC)

Yes, I'm not mentioning anything in specific, just discussing 'tools' and 'NICs' in general without mentioning any specific tool project or NIC vendor. I also agree that the overly technical details aren't needed. We're writing for a general audience here, not an audience of network engineers. MrOllie (talk) 02:53, 25 October 2021 (UTC)

I added a reference to the other tool (MrOllie included the reference to one) without mentioning the name of either of the tools. -Macmobile (talk) 05:41, 28 October 2021 (UTC)

One last time, I highly recommend adding the following - which is concise, and I don't really see how this may hurt the article but it definitely helps people who might be interested to get into the field - but if both of you insist I'm fine to let go. I'm done with the discussion - and feel free to clear the talk page at your will. -Macmobile (talk) 05:41, 28 October 2021 (UTC)

The CSI measures the channel matrices from a variety of OFDM subcarriers that can describe the frequency responses of the 20MHz/40Mhz wireless channels used in 802.11n. The primary motivation and uses were for wireless channel estimation, network troubleshooting, developing Multi-User MIMO systems, and understanding channel fading in Wi-Fi networks. — Preceding unsigned comment added by Macmobile (talk • contribs)

I reused a citation that was already present in the article. There's no need for a redundant citation. Is that what this is really about? Are you associated with that tool, paper or its authors somehow? MrOllie (talk) 12:00, 28 October 2021 (UTC)

I worked in the area of wifi sensing and used to be a user of both tools - and I believe I have much more experience and expertise than you to talk about its "relevance" or "redundancy" to wifi sensing. However I'm tired of the argument and would let go. -Macmobile (talk) 07:24, 29 October 2021 (UTC)

References