Ordinary WiFi can now identify people with near perfect accuracy

We used to recommend people to run the newest stuff possible, but we came to a point that maybe it's better for us to keep with older tech for a good while

From what I've just read, the tech doesn't seem ready to identify people yet. It can supposedly detect hand gestures, but facial recognition I seriously doubt. But that's probably just a matter of improving the tech. See this article for more info.

That's connection history. CSI motion detection software storing information it collects would be entirely independent of that. How much it saves and for how long would depend on the size of the router's memory.

I'm already envious 🙂‍↕️

They're cucks. They're into that.

Tadaa https://github.com/francescopace/espectre

Yeah, if this shit has to exist, at least let me use it for presence detection in Home Assistant without having to buy separate sensors or something!

If you're technical you might like enjoy this article that explains how the tracking works. Basically the router can perform math on the interference created by objects moving around the room. It seems like this would have to be part of the router firmware, which doesn't sound like a standard feature. But if it is, the fix would be to install modified firmware with that function disabled. The smoking gun will be if somebody gets into DMCA trouble for doing this.

Or an open source hardware device that changes your "wifi signature" randomly.

Opensource tech to do the same thing has been in the hands of the people for a long time. This is just a different way of doing it without motion sensors.

app to manage router

This shit was a pain in the ass and now learning about this makes me feel even more pissed off as a customer

Damn. Put a faraday cage around the router and plug in your own router to a LAN port.

FWIW I was able to use my own router when I set up with Xfinity recently

That's because Xfinity offers motion sensing as a feature, which requires this tech in the router. Presumably it's configurable and costs extra to turn on.

Comcast is why I have starlink

Gimme cat boobs.

wearing a smartwatch that constantly outputs an identifier.

It is easier to just give up and submit, I'll grant you that.

Huh? No cats on mine , weird.

Stealth Bomber Jacket.

Don’t give Musk the idea of the CyberShirt.

Funnily enough, indoors, this would probably make you more visible as the only area with no reflections. Stealth works outdoors because the sky does not have a radar return.

My understanding is that this catches disruptions between devices and router. I don’t think this would work. I would say you should instead sell a “bracelet” with “ancient Himalayan Salt” embedded into the silicone to absorb and cancel the tracking. It would probably sell a ton! Obviously wouldn’t work but $!

You can buy faraday bag cloth. It’s expensive.

That's using CSI though. The article said the researches specifically did not utilize CSI.

But regarding CSI: I evaluated that as a small part of my Master's thesis and it worked pretty OK for motion detection but not for classifying other activities, at least not on a SISO link. For more complex stuff you would need both a MIMO access point (router) and device (e.g. phone). Also, you need to constantly transmit messages to get up-to-date CSI, which is not great for power consumption as well as cluttering the communication channel. There are some other constraints, especially regarding noise. E.g. I managed to completely destroy the CSI spectrogram by turning on a microwave oven. There is 802.11bf in development, which is supposed to standardize this, because currently using CSI is pretty much a "hack", as it is not intended for sensing. Once this is widely adopted, I would start being worried, but not right now.

This is from my thesis:

When an inert object like a person

Say waaaa?

Right? Im pretty sure this is a few years stale and already incorporated in some isps routers

https://www.xfinity.com/hub/smart-home/wifi-motion

https://www.originwirelessai.com/isps-can-do-more-with-wifi-sensing/

https://www.technologyreview.com/2024/02/27/1088154/wifi-sensing-tracking-movements/

https://github.com/ruvnet/RuView

The first time I heard about this was in 2013 and, in 2019, I had a local government management class where wifi sensing in busy downtown areas and stadiums was discussed as a plus side to municipal wifi installations. In the latter case it was described as being available not too far in the future.

Like a submarine.

I'm not sure what you think an "example" would look like. It's not taking a photo of you, it's measuring what's distinctive about the way you personally mess up radio signals and how it differs from how other people mess them up. Internally it's just a ton of numbers.

a real life example? you mean like a photo of a person next to a router?

and how do you protect yourself against the neighbors devices, especially in a densely populated building

I mean, they kinda already were.

That means "yesterday's spy tech" that now they will leak to public, because they have a way better way.
I'm not sure of the current state of my tinfoil hat.

That would prevent cell signals from inside, making it harder to (e.g.) call the fire department, or an ambulance.

Kinda defeats the purpose of wifi

I would not be wild about that. I use my wifi outside.

I'm wondering if some types of crystal can absorb or distort the signals enough to make you harder to identify. I mean I guess you'd be "that guy with the crystals in his pocket" and would be easy to correlate to CCTV flock, but if everyone does it it could create a layer of obscuration.

Especially things with piezoelectric and/or electromagnetic effects. I'm thinking mainly along the lines of quartz, tourmaline, pyrite, tiger's eye (hematite/jasper), shungite, etc.

It would be worth experimenting with if anyone tries that open source software another commenter linked...

Why not both?

I can tell you as someone who read the papers on very early deepfakes and AI video generation with amazement followed by dread, this is going to be feasible on a large scale in a short period of time. Researchers do stuff on an absolute shoestring budget usually, it's incomparable to what large companies and governments have at their disposal. There are already consumer products that were able to become fairly precise motion sensors with just a firmware update. Next gen devices will be built with motion fingerprinting in mind, I can almost guarantee it.

It gets more accurate with more access points, too. So corporate and education settings will be the easy places for this to get implemented.

It's also only possible because the information they used (BFI) is unencrypted.

It's a start. It may take time to make it work for "everyday" use, but if it's possible now, it can be done better in the future.

What's an AP?

They're not all sending at the same time. Worst case they just block themselves and each other with their backoff logic and then none of them sends anything at all.

are you sure it works that way?

A paper from around a decade ago talked about using WiFi to identify key strokes so with large data models we have today I would assume they could get pretty good fidelity on a person. Maybe not enough for “beyond a reasonable doubt” but probably enough where your WiFi company is selling your data on what you do at home

Height and build alone would narrow it down significantly.

The wavelength of a 5 Ghz wifi signal (the highest frequency in common use for wifi) is a little under 6 cm. So as a crude measure, it's not going to resolve spatial features much smaller than that with much reliability.

I think as a person moves about the waves would get a clearer picture.

You can initiate identification post visual contact or before loss of visual contact. As long as tracking is good the identification info can be propagated

Most people in a “technology” forum completely understand this and yet are also still capable of reading and understanding TFA.

Yes, but 90% of ISP supplied modems around the world are modem + router + WiFi access point with a unified firmware.

You also can't take the antennas off of those and they are required in order to receive internet.

Yes you can use your own router (I have a Unifi cloud gateway ultra myself and one access point in the middle of the house), but that doesn't mean that disabling the WiFi on the ISP web-software bullshit actually disables the WiFi and doesn't just hide the SSID and make it un-connectable and still use it for this kind of thing and identifying nearby devices.

It also doesn't mean that all the routers themselves like my Unifi aren't using the access points to do the exact same thing (or will in the future). The only way you can actually control that is with openwrt or similar.

You being wired doesn't stop WiFi seeing you.

Sure as well is awkward for a mobile phone.

I've seen YouTube videos of people able to record the image of the the vibration of a potato chip bag through a window to recreate the audio from the room.

There is now a free home assistant plugin to implement it at home lol. Crazy shit. Imagine what's possible with classified tech.

https://github.com/francescopace/espectre

I don't think a person has to carry anything. The tracking is based on measuring the interference a moving person (or a dog) creates between the router and a connected device like a range extender or a networked printer.

Yes, it wouldn't really.

Right now the way this works is that a human body absorbs a certain about it wifi signal, so of the signal strength in a room dips and comes back up, someone walked through the room, for example. Couple this with what IPs and MAC addresses the router is connecting to, and Verizon can tell "human with laptop," or "human watching TV." So just "human body" or dog/cat are what it can detect. Verizon does try and sell this as a feature, as in a shit security feature.

So for cell towers, they're too far from people in an already chaotic environment to really be useful. Trees, cars, and a million other things can throw off trying to detect already minute changes in signal density. Not to mention that the signals from cell towers are much stronger, so harder to detect the changes.

I think the main advantage with the wifi-based approaches is that they are usually used in a relatively static/calm indoor environment with a stable channel response and your motions are disturbing that, compared to a quickly changing outdoor environment (e.g. a city) where it would be much harder to distinguish individuals. Also, you are typically closer to the access points, making the power/SNR higher. Regarding mobile communication though, the trend is towards higher frequencies and smaller cell sizes which also give greater spatial resolution (and higher power) and some funky near-field effects can be used to get beam forming on crack: https://arxiv.org/abs/2407.10147 So perhaps it could work even better, wouldn't be surprised

I imagine resolution decreases with range

Edit: resolution not revolution

There's no need, they can use triangulation since you're almost always near your phone

Despite what others have said. It could in theory. But could it work with ordinary cell towers today, probably not. I base this on the accuracy of current location tracking by cell towers. They still use triangulation from my understanding, and aren't highly accurate at that. The space your phone could be in is large enough for many people to be. So the granularity just isn't there.
This is probably because of the large range they cover compared to the power levels they use. But in theory if the density of towers were higher, and the power levels were increased, they could probably do it in at least some locations with the perfect conditions.

I know of ESPresense, but that only tracks your phone, not your body..

Assert dominance

Height and body mass

To confirm; turn off your adblocker and see if you have any relevant ads.

I clearly need watch more Batman.

Arguably that's a bit difference because to do that you have to explicitly do it (room setup) and you view the result (visual preview with semi-transparent triangles over your place). You can also read the ToS and I believe in some case specify if you allow the information to be sent back to the Meta. I'm not saying it's OK, only that it's explicit and it's part of the "normal" usage of the device.

The AP is what does the tracking. If you build your own, and only you can access it, then it can’t watch you. VPNs don’t solve any of this, it’s not about a device being encrypted or not.

I'm not sure about wifi but cell jamming has been heavily criminalized and tracked

I mentioned this in a another reply, but you just need a microwave oven basically :)

Particularly annoying snoopers sounds like some people in my nieghborhood

I need to move 🚚

You don't need to build your own router. All you need to do is flash open source firmware like OpenWRT on a supported router to prevent it from spying on you.

Nothing. Get a mini pc with 2 or more ethernet ports and install openwrt or opnsense or maybe others I haven't tried, get a standalone wifi AP so if you ever want to upgrade anything you can do it separately, get a network switch to connect additional wired computers or wifi APs.

Then get in all the homelab groups and find the coolest stuff you also want and start making the home network made by you and for you.

If you already have a wifi router that can be flashed with custom firmware you can also use that to start out and use it as a wifi access point for a more powerful pc router later

Knowing how.

All routers are just Linux computers managing slightly better than average network cards. (Home routers, obviously. Beefy carrier hardware is different).

A basic setup is essentially installing Linux and then running a handful of commands for packet forwarding. The figuring out how to do it without wifi crashing will take longer because that software is wonky.

You will be identified by the unique features of the said mask. And if you happen to move in public, by your gait properties.