Wi-Fi signals allowed the smartphone to ‘see’ the person behind the wall

American engineers have developed a method for using Wi-Fi-enabled devices in the room and an outdoor receiver to track the presence of people in the room. The created algorithm first calculates the location of the transmitters in the room from signals, and then determines the presence of a moving person in it from the reflections of the signals, the developers say in an article published on arXiv.org.
 
Today, wireless data transmission according to the Wi-Fi standard has become widespread – wireless routers are installed in many apartments and offices, and smartphones, tablets and other devices are connected to them. Because of this, some experts in the field of communication and computer security investigated Wi-Fi as a way to obtain information about the movements of people and other objects. For example, last year a group of researchers learned how to “highlight” a room using a Wi-Fi enabled device and collect information about objects in it using a receiver outside. However, earlier such studies meant that to obtain data, you must either place the Wi-Fi transmitter inside the room and know its location, or use powerful equipment outside.
 
Haitao Zheng from the University of Chicago and his colleagues have shown in practice that a smartphone capable of receiving a Wi-Fi signal is enough to obtain data on the actions of people inside the premises. The attack described by the developers implies the use of routers, smart speakers and other devices located stationary inside the room for “highlighting”. However, with such a scheme, a technological problem arises – it is difficult to calculate the reflection of Wi-Fi signals from the target, if the location of the signal sources is unknown.
 
Because of this, the method proposed by the engineers consists of two stages: localization of signal sources and direct observation. They created a smartphone application that simultaneously records the received signal power for each Wi-Fi device in the room, as well as the movement of the smartphone based on accelerometer data. The attacker just needs to walk past the premises with the included application to collect enough data.After that, the smartphone must be placed motionless near the room or several rooms. Since the program now knows the static positions of the indoor transmitters and the receiver outside, data on its movements can be calculated from changes in signals due to reflections from a person.
 
The developers conducted several experiments to test the method, conducted in 11 different rooms using a variety of devices of various types, including smart speakers and cameras, routers, smartphones and others. Tests have shown that the effectiveness of the method strongly depends on the number of Wi-Fi devices operating as radiators. For example, for four devices, the frequency of false negatives (for which the device does not detect the person who is actually in the room) was 0.19 percent, and the frequency of false positives was 27.49 percent.Last year, another group of American developers introduced the technology of three-dimensional scanning of objects hidden behind walls, which is also based on Wi-Fi. The system consists of two drones, one of which emits a Wi-Fi signal, and the second receives it. Since the movements of the drones are synchronized, a three-dimensional model of the hidden object can be compiled by attenuating the signal.

---