Researchers at a German institution have developed a technique that uses standard WiFi signals to identify individuals with near-perfect accuracy, without requiring people to carry devices or have their phones powered on.
The method works by analyzing how radio waves reflect and bounce throughout an enclosed space. WiFi transmitters emit signals that scatter off walls, furniture, and human bodies before returning to receivers. Each person creates a unique pattern of reflections based on their body shape, size, and movement. Machine learning algorithms trained on these wireless signatures can recognize specific individuals reliably.
The technology essentially transforms WiFi infrastructure into a passive sensing system. Unlike traditional surveillance cameras or motion detectors, it requires no additional hardware beyond existing routers. The approach raises immediate privacy concerns because WiFi networks permeate most buildings, and people generally assume these signals cannot be weaponized for identification purposes.
The researchers demonstrated the system's capability across multiple test scenarios, achieving accuracy rates approaching 100 percent in controlled settings. The technique works through walls and solid objects, meaning recognition occurs even when subjects are in different rooms from the WiFi equipment.
The work builds on years of research into "WiFi sensing" or "passive radar" using wireless signals. Previous studies showed WiFi could detect human presence and track movement. This advance moves the technology toward reliable personal identification, a significant leap in capability with substantial implications for privacy.
The implications extend beyond academic interest. Any entity with WiFi access in a building could theoretically identify occupants without their knowledge or consent. The technology requires no special equipment from targets, no active participation, and produces no obvious signs of surveillance.
However, real-world applications face practical constraints. Performance may degrade in complex environments with multiple people, moving objects, and network interference. The algorithms appear to require training data specific to target individuals beforehand. Scaling the technology to crowded public spaces remains uncertain.
The researchers have not specified their institution or publication venue. Their