Researchers at the Karlsruhe Institute of Technology have demonstrated that standard Wi-Fi routers can be used to identify people inside a room without cameras or wearable devices. Their method, called BFId, analyzes unencrypted beamforming data that Wi-Fi devices routinely transmit. It identifies a person by how their movements alter the radio signal.
A key advantage of this technology is that it doesn't require connecting to the target network or any special hardware. All it needs is a device with a Wi-Fi adapter in monitoring mode, passively collecting service data. The person being tracked doesn't even need to carry a smartphone, smartwatch, or other wireless gadget.
Previous research in this area often relied on channel state information (CSI), but obtaining that data usually requires modified firmware and compatible network cards. BFId takes a different approach, using beamforming feedback information instead. This data first appeared with Wi-Fi 5 and helps routers direct signals more precisely to connected devices. The catch is that this information is transmitted unencrypted at the MAC level, making it easy to intercept.
In tests with 197 participants, the system achieved an accuracy of up to 99.5%. On the same test set, BFId significantly outperformed CSI-based methods: 99.5% versus 82.4%. The researchers attribute this to the fact that compressed beamforming data filters out some noise and provides more features for analyzing human movement.
Simple countermeasures proved ineffective. For instance, reducing the frequency of beamforming reports barely affected recognition accuracy. Encrypting the data could solve the issue, but that would require changes to Wi-Fi standards and potentially break compatibility with older devices.
The researchers caution that while Wi-Fi sensing could be useful for presence detection, room monitoring, and smart home applications, it also raises serious privacy concerns. This is especially relevant given the 802.11bf standard, which formalizes such capabilities. According to the team, protective measures need to be put in place before Wi-Fi sensing becomes widespread.