Wi-Fi信号可“隔墙识人” 警惕无线网络存在隐形监控威胁

Core Insights - A new study reveals a concerning surveillance threat where attackers can identify individuals nearby using existing Wi-Fi signals, even without the use of Wi-Fi or when devices are turned off [1][2] - The research indicates that attackers can passively listen to changes in radio wave propagation caused by human bodies, allowing for identification without any special hardware [1][2] Group 1: Technology and Methodology - The technique relies on the beamforming feedback signals sent by devices to Wi-Fi routers, which are unencrypted and can be intercepted by malicious listeners [1] - By analyzing how these signals change due to human movement and posture, machine learning models can be used to construct movement characteristics and identify specific individuals [1][2] Group 2: Implications and Risks - Every Wi-Fi router could potentially act as an "observer," enabling long-term tracking of individuals even if they have never connected to the network [2] - The method is less intrusive compared to previous technologies that relied on expensive sensors, making it harder to detect and not limited by physical line of sight [2] - In an experiment with 197 participants, the identity recognition accuracy approached 100% within seconds of signal analysis, highlighting both the technology's potential and significant privacy risks [2]