Identify People Through Walls Using Only WiFi

WiFi signals are all around us in many areas, and on some level we know that our body presence affects how the signal travels. Researchers at the UC Santa Barbara, however, have taken this concept several steps further, using WiFi signals to sense people through walls with an overall accuracy of 84 percent.

The system, which they refer to as XModal-ID (pronounced Cross-Modal-ID), uses a WiFi transmitter and receiver pair situated just outside of a wall to sense a moving person, then compares how he or she moves to what is expected based on video footage.

XModal-ID translates the video into a 3D mesh that describes the outer surface of a human body as a function of time, and utilizes an electromagnetic wave approximation to simulate the RF signal that would have been generated by the person if walking in the WiFi area. Finally, the data is compared to the gait of the person seen via the WiFi setup and the simulated RF signal to see whether a particular person is moving about.

One proposed application of the system would be if police have video footage of a robbery. If they suspect that the perpetrator is hiding inside a building, they could accurately pick out whether that person is present based on this type of WiFi configuration. It’s a truly amazing concept, though one that could perhaps help justify your paranoia!

More information can be found on the team's website.