SenShield: Preserving User Privacy Against Passive WiFi Sensing

• Applicant: Professor Dr.-Ing. Matthias Hollick, since 4/2024
• Subject Area: Security and Dependability, Operating-, Communication- and Distributed Systems
• Term: from 2020 to 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 447586980

Within the past decade, researchers have shown that WiFi signals can be used for a variety of sensing applications ranging from human motions to emotions. Lately, device-free WiFi sensing has gained traction since it does not require the target to carry a device. Although device-free, these solutions proved highly accurate in tracking movement and trajectory of the users on decimeter-level as well as recognizing their activities (e.g., running, walking) and fine- grained gestures (lips, fingers). Indeed, WiFi sensing is a convenient method for such tasks, given the abundance of WiFi- enabled devices around us. From a technical point of view, WiFi sensing boils down to extracting the impact of human body motion on the signal propagation, which is conceptually similar to bistatic radar. Recently, a few startups (e.g., Totemic, Emerald) ventured into this area, branding their products as non-invasive and privacy-preserving health monitoring devices, as opposed to video and on-body sensors. However, the maturity of the technology, its practicality, and the abundance of WiFi devices pose significant risks to user privacy. In its simplest form, a burglar can identify empty houses or profile them based on the number of occupants and even the health of the occupants through monitoring their moving speed or heartbeat. The more significant privacy risk is that, unlike radars, the adversary can leverage WiFi sensing techniques passively (i.e., without transmitting any signal), hence going undetected. Furthermore, WiFi device manufacturers and government agencies can potentially conduct mass surveillance with unnoticeable firmware manipulation. Unfortunately, even in the presence of such risks, we cannot defy the use of WiFi devices since they are already an integral part of our lives at home and work. Although WiFi sensing is conceptually similar to radar, we cannot leverage classic radar jamming techniques because they interfere with legitimate communication over WiFi. In SenShield, we devise fundamental solutions to minimize the concerns raised by the recent advent of WiFi sensing techniques. More specifically, we aim at thwarting passive WiFi sensing possibilities where the adversary does not transmit any signal and tries to obtain all necessary information via eavesdropping on ongoing legitimate WiFi communication.

• DFG Programme: Research Grants
• Ehemaliger Antragsteller: Arash Asadi, Ph.D., until 3/2024