Swedish researchers use Wi-Fi signals and AI to detect hidden firearms
Scientists at Sweden's Defense Research Agency have built an experimental system that identifies concealed handguns by analyzing how Wi-Fi signals scatter around objects, sidestepping the need for cameras or metal detectors.
Researchers working at the Swedish Defense Research Agency (FOI) have created an artificial intelligence system capable of recognizing concealed handguns through Wi-Fi signal analysis, presenting a less intrusive option compared to traditional surveillance cameras and metal detection equipment. The work, presented in a paper uploaded to arXiv on October 13, describes a tool named WiFireNet that processes channel state information—the data produced when objects alter the propagation of Wi-Fi signals in their vicinity. The complete setup includes a Wi-Fi transmitter, a Wi-Fi receiver, and the WiFireNet AI model.
How the system works
When Wi-Fi signals travel through space and strike an object, they pass through it and scatter in patterns determined by the object's physical characteristics—its size, angle, and composition. Researchers can extract and examine this channel state information to determine what object caused the signal distortion. Earlier work by security researchers demonstrated that standard Wi-Fi equipment can identify people with 99.5% accuracy using comparable signal analysis methods.
Current limitations
The current prototype version of the system cannot determine the exact position of a weapon on a person's body, representing a significant constraint on its practical deployment.