SoK: Inference Attacks and Defenses in Human-Centered Wireless Sensing

Human-centered wireless sensing aims to understand the fine-grained environment and activities of a human using the diverse wireless signals around her. The wireless sensing community has demonstrated the superiority of such techniques in many applications such as smart homes, human-computer interactions, and smart cities. Like many other technologies, wireless sensing is also a double-edged sword. While the sensed information about a human can be used for many good purposes such as enhancing life quality, an adversary can also abuse it to steal private information about the human (e.g., location, living habits, and behavioral biometric characteristics). However, the literature lacks a systematic understanding of the privacy vulnerabilities of wireless sensing and the defenses against them. In this work, we aim to bridge this gap. First, we propose a framework to systematize wireless sensing-based inference attacks. Our framework consists of three key steps: deploying a sniffing device, sniffing wireless signals, and inferring private information. Our framework can be used to guide the design of new inference attacks since different attacks can instantiate these three steps differently. Second, we propose a defense-in-depth framework to systematize defenses against such inference attacks. The prevention component of our framework aims to prevent inference attacks via obfuscating the wireless signals around a human, while the detection component aims to detect and respond to attacks. Third, based on our attack and defense frameworks, we identify gaps in the existing literature and discuss future research directions.