What if they could see YOU using WiFi?

Proton，Proton 官方博客，加密邮件、VPN、云盘与密码管理器等隐私保护生态。关注

WiFi Sensing

Cameras may be ubiquitous, but they’re also often quite easy to spot; a camera will also need light and some kind of lens in order to operate. A router, on the other hand, requires none of this… and nobody at the moment thinks of it as a camera.

Researchers at Karlsruhe Institute of Technology have just shown it can be one, given the right setup. A presentation at the ACM Conference on Computer and Communications Security (CCS) in Taipei last year surfaced a new worry for the privacy-conscious (or for anyone, for that matter): WiFi Sensing.

The presented paper showed that, across 197 people, individuals could be identified with near-perfect accuracy. This finding was robust across different angles and it didn’t matter how each person walked. Just so these findings are not misinterpreted, these were known individuals. An entry into a database could be reliably matched with an individual, we aren’t talking about ID-ing strangers on the fly.

Radio waves here were performing the same job that light waves do in a camera. This is not “your phone is tracking you” as it works regardless of if you’re carrying one. All of this is done using the magic of radio waves. Let’s dig in…

How It Actually Works

In order to aim signals efficiently, WiFi devices constantly report back to their router about how the radio channel itself looks; the information transmitted here is called beamforming feedback information (BFI). This transmission, which became consistently present with WiFi 5 (IEEE 802.11ac), is sent unencrypted (because no-one ever thought that it could be sensitive,) so it is readable to anyone in range.

Earlier versions of WiFi Sensing leaned on channel state information (CSI) or using specialist kit like LIDAR, which means that in order to do this same mapping of people you would have to have a considerable amount of equipment. Using BFI means you can sidestep this effort or expense, it is a compressed and already-broadcast byproduct of normal WiFi usage… and a standard device is all you need to collect it.

You would expect BFI to be worse than CSI for sensing, as it is a lossy, compressed version of the same thing; this being said, the compression strips out noise, and the ease of collecting these signals means that it is a trivial thing to grab it from several different points at once. We are unfortunately dealing here with an easier attack that leverages less data.

And just to be clear here, this is not reading your traffic, getting into your network, or following a MAC address. Your body here is a fingerprint which is built from antenna signals. Pretty shocking stuff.

Why This Matters

Obviously there are considerably-easier ways of watching people. A pretty and worryingly-large quantity of people think that the privacy cost of installing an Amazon CCTV camera on the front of your house doesn’t outweigh getting parcels delivered reliably. This isn’t the end of privacy, it’s just another thing that we should be aware of as we attempt to minimize how much we’re being watched.

What does make it frightening, and here I’m cribbing from the researchers themselves, is that the ubiquity of WiFi means that we could’ve accidentally created a near-comprehensive radio-wave panopticon that is completely invisible. Once again, cameras normally can be picked out; routers are in our homes with this latent ability to become little snitches (and not the good macOS-software kind).

If you walk past the same cafe regularly, without knowing you could be monitored by these unseen radio signals. The researchers themselves here point to the vulnerability of protesters and dissidents in authoritarian states, where visible cameras could be superseded by this pervasive, and heretofore unknown, tracker.

The History of WiFi Sensing

The evolution of this process has been quite remarkable: from WiFi that reads through walls, to RF-Pose-style mapping of bodies using radio, to CSI activity sensing. A body of research where every change has required either more kit or more equipment than the previous one… until now.

In fact, this process needs less than any of the others. Sensing has been moving away from labs with specialized hardware and toward basically anything that speaks in the language of modern WiFi. This capability wasn’t one that just appeared overnight; rather, it got cheaper, more passive, until it reached gear that you or I could reasonably set up.

Much like other articles I’ve published here, this is the same pattern as TVs, as wearables… features which were added for a different reason create some kind of data-stream which, in turn, becomes a plausible surveillance nightmare. Nobody built BFI in order to identify people, it was to optimize signal, but it does do this, and that’s worrying.

What Actually Helps

You can’t meaningfully stop your body from reflecting radio waves (I’m going to seem stupid when someone comes up with a bodysuit Faraday cage) and turning off devices is not going to help, because other nearby active WiFi devices will suffice.

Unfortunately, this issue is one of the ones where the solution is very much upstream. Privacy protections need to be built into upcoming standards before this scales. Encrypt or restrict BFI, and design the standard so that this side effect isn’t there, and the problem kind of goes away.

Fewer always-on wireless devices means a quieter radio environment, but really this is a thing that needs to be pushed for: baking privacy into the standards that inform how our interconnected world is built.

The Bigger Picture

For ages one of the easiest ways to dodge a whole host of different forms of tracking was just to leave your phone at home, cover your camera, chuck it into some kind of signal-blocking pocket. This paper shows that you can have a form of tracking done to you without the need for you carrying a device, because the signal is created by what is reflected off of you. The infrastructure that we have filled our homes, streets, offices, and public spaces with can be repurposed in order to watch the people that are moving through these spaces.

We opened the door to these little broadcast devices when they weren’t cameras; now it’s time to push back against their latent potential to surveil.

Sources

添加评论
点赞收藏
点踩分享查看原文
评论
?
参与讨论