Ordinary routers can recognize a person by reflecting a radio signal from the body.
A person enters a cafe without a smartphone, but a working Wi-Fi network can still betray his identity. Researchers at the Karlsruhe Institute of Technology have shown that radio signals of conventional wireless devices allow to recognize nearby people, and the attack does not require cameras, special sensors or victim equipment.
The BFid method uses Beamforming Feedback Information, or BFI. Wi-Fi devices regularly transmit the router such information to the network better. BFI packages are not encrypted, so the nearby observer can record them using standard Wi-Fi equipment.
Radio waves change when reflected from walls, furniture and the human body. The system analyzes the obtained changes and forms a kind of radio images of a person from different angles. After training, the machine learning model learns a familiar participant in the experiment in a few seconds.
Wearing a smartphone, laptop or smart watch for such a surveillance is not required. The turned off phone also does not protect a person if other devices connected to the wireless network work nearby. According to Professor Torsten Strufe from the KATEL Institute at KIT, the principle resembles a camera, only instead of light, the system uses the propagation of radio waves.
The researchers tested the method on 197 participants. In the experiment, the system defined the personality with almost 100 percent accuracy, and the result was not dependent on the direction of observation and the manner of walking. The authors of the work warn that the laboratory result turns the usual networks into a potential device for hidden observation.
If you regularly pass by a cafe with Wi-Fi, you may be imperceptibly identified, and later learn again, for example, government agencies or companies,” said KASTEL researcher Julian Todt. His colleague Felix Morsbach admits that it is still easier for the security services and attackers to hack surveillance cameras or video intercom. The danger of Wi-Fi is different: wireless networks surround people almost everywhere and do not cause suspicion.
The authors consider the use of technology to be particularly dangerous for spying on protesters and citizens in countries with repressive regimes. The team called for the integration of protective mechanisms into the developed IEEE 802.11bf standard associated with the detection of objects and people by Wi-Fi.
A person enters a cafe without a smartphone, but a working Wi-Fi network can still betray his identity. Researchers at the Karlsruhe Institute of Technology have shown that radio signals of conventional wireless devices allow to recognize nearby people, and the attack does not require cameras, special sensors or victim equipment.
The BFid method uses Beamforming Feedback Information, or BFI. Wi-Fi devices regularly transmit the router such information to the network better. BFI packages are not encrypted, so the nearby observer can record them using standard Wi-Fi equipment.
Radio waves change when reflected from walls, furniture and the human body. The system analyzes the obtained changes and forms a kind of radio images of a person from different angles. After training, the machine learning model learns a familiar participant in the experiment in a few seconds.
Wearing a smartphone, laptop or smart watch for such a surveillance is not required. The turned off phone also does not protect a person if other devices connected to the wireless network work nearby. According to Professor Torsten Strufe from the KATEL Institute at KIT, the principle resembles a camera, only instead of light, the system uses the propagation of radio waves.
The researchers tested the method on 197 participants. In the experiment, the system defined the personality with almost 100 percent accuracy, and the result was not dependent on the direction of observation and the manner of walking. The authors of the work warn that the laboratory result turns the usual networks into a potential device for hidden observation.
If you regularly pass by a cafe with Wi-Fi, you may be imperceptibly identified, and later learn again, for example, government agencies or companies,” said KASTEL researcher Julian Todt. His colleague Felix Morsbach admits that it is still easier for the security services and attackers to hack surveillance cameras or video intercom. The danger of Wi-Fi is different: wireless networks surround people almost everywhere and do not cause suspicion.
The authors consider the use of technology to be particularly dangerous for spying on protesters and citizens in countries with repressive regimes. The team called for the integration of protective mechanisms into the developed IEEE 802.11bf standard associated with the detection of objects and people by Wi-Fi.
