As time passes, my deep-seated desire to not be perceived only grows. Unfortunately, becoming truly imperceptible both online and off clashes with a number of my ambitions—I can't keep writing if there's no audience, right? Well, going off the grid and becoming a cryptid in the woods looks like it's fully off the cards for me now, thanks to recent Wi-findings from researchers in Italy.
A team out of La Sapienza University of Rome has developed 'Who-Fi,' a system that can track individuals with an alarming level of precision by monitoring how their physical body interacts with Wi-Fi signals (via PCWorld). Based on how one's body interferes with this ubiquitous network of signals, the researchers can effectively 'fingerprint' folks and track them with up to 95.5% accuracy. Is anyone else feeling kind of claustrophobic within their flesh prison today, or is that just me?
The research paper goes into more depth, explaining how various biomarkers are "extracted from Channel State Information (CSI) and processed through a modular Deep Neural Network (DNN) featuring a Transformer-based encoder."
The research team pitches their 'Who-Fi' method as an alternative to traditional visual tracking over video feed, which can be made more difficult by "poor lighting, occlusion, and suboptimal angles." However, my skin really started to prickle when I read that "Wi-Fi signals offer several advantages over camera-based approaches," such as possessing the ability to "penetrate walls."
What is broadly known as Wi-Fi Sensing has already been leveraged to a variety of much less anxiety-inspiring ends, such as waking up and putting your PC to sleep. Over the years, a number of research teams have also used Wi-Fi Sensing as a non-invasive alternative to monitoring patient respiration—you can watch Intel Labs researcher Max Pinaroc talk through one such breath-detection demo here. But this time it's difficult to overlook the surveillance implications.
For one, this method doesn't rely on placing someone based on their phone's position in relation to cell network towers. If you're going to, oh I don't know, attend a perfectly innocuous gathering to show solidarity around a shared cause, you can't just leave your body at home like you can a phone.
Thankfully, this technology has yet to be used in such a real-world scenario. Even though the research team didn't need masses of specialised hardware to pull off their Wi-Fi Sensing project—apparently recording their data using only "two TP-Link N750 routers"—it would be premature to get too paranoid just yet.
Digging into that 95.5% accuracy rate I mentioned earlier, it turns out this relates to what is actually a pretty small sample size of "14 different subjects." The paper explains that "60 samples were collected [from each subject] while they were performing a short walk inside the designated test area," with 'Who-Fi' able to accurately identify each person through a variety of simple appearance changes, including "wearing only a T-shirt, a T-shirt and a coat, and a T-shirt, coat, and backpack."
So, 'Who-Fi' is hardly a cyberpunk Eye of Sauron staring into your very soul just yet, instead sitting mostly within the realm of 'ain't it neat'—for the time being, at least.
1. Best overall:
HP Omen 35L
2. Best budget:
Lenovo Legion Tower 5i
3. Best compact:
Velocity Micro Raptor ES40
4. Alienware:
Alienware Aurora
5. Best mini PC:
Minisforum AtomMan G7 PT
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
