Who-Fi is an innovative technology empowered by artificial intelligence (AI) that can identify and monitor individuals without the need for visual cues. This experimental system, while still undergoing real-world testing, has the potential to transform ordinary Wi-Fi signals into a powerful biometric scanner. This scanner not only tracks the movements and positions of individuals but also recognizes their distinct biometric characteristics.
Understanding the Who-Fi Technology
As detailed in a paper on arXiv, standard 2.4GHz Wi-Fi signals serve as the basis for identifying and monitoring individuals, playing crucial roles in identity verification and surveillance. However, the technology also introduces new concerns regarding digital privacy and security.
The Who-Fi system utilizes a blend of Wi-Fi signals and a transformer-based neural network, also known as a large language model (LLM). This LLM interprets a concept called ‘channel state information’ (CSI), tracking variations in Wi-Fi signal strength and phase as they interact with an individual’s body. This analysis mirrors the methods of radar and sonar systems.
When a person is in proximity to a Wi-Fi signal, the signal’s path distortion creates a unique pattern. This pattern rivals the accuracy of traditional biometric markers such as fingerprints, facial features, and retina structure. By recognizing this pattern, the Who-Fi system can attribute it to specific individuals.
Once the system is trained on these biometric signatures, it not only tracks individual movements but can also identify them upon re-entering the network zone after a prolonged absence. It also captures body movements and can decipher sign language. A notable feature of the system is its operation without visual or auditory sensors, such as cameras or microphones.
Efficiency and Capabilities
The Who-Fi system comprises a single-antenna transmitter and a three-antenna receiver, resulting in a cost-effective deployment. Researchers have demonstrated that even when a target individual is behind a wall and moving at a typical pace, Who-Fi achieves an impressive 95.5 percent precision.
Moreover, the system’s accuracy remains consistent even if individuals change attire or carry items like backpacks. Remarkably, a single system can monitor and identify up to nine individuals simultaneously.
Who-Fi boasts high evasion capabilities, making it challenging to detect through surveillance technologies. This stealth is achieved without specialized hardware or emission patterns like infrared, radar, or visible light, as the system engages in passive radio frequency (RF) sensing to maintain a discreet profile.
