Deception in Wireless Communications: Benefits and Perils

Abstract

Wireless signals, ubiquitous and capable of penetrating obstacles, are increasingly exploited to infer personal information such as vital signs, human activities, and crowd counts. This surge in wireless sensing techniques benefits society but also poses significant security threats. Adversaries could potentially use these signals for wireless inference attacks, determining a target's activities with little regard for privacy or security. This dissertation mainly includes two studies on deception in existing wireless networks and systems. On one hand, it proposes proactive measures to send deceptive signals to mislead eavesdroppers, thus convincing them their attempts are successful when, in fact, they are failing. On the other hand, it identifies vulnerabilities in wireless liveness detection systems, where an attacker can use such wireless deceptive techniques to compromise these systems. The first study proposes proactive defenses against all existing Channel State Information (CSI)-based vital signs and crowd counting inference attacks by establishing ambush locations and then transmitting deceptive signals designed to mimic vital signs or person count, albeit falsely, thereby protecting the user's true vital signs or person count data. Experimental results on software-defined radio (SDR) platforms demonstrate that these defensive strategies can effectively mislead eavesdroppers, who then receive falsified breathing rates or person counts. Conversely, the second study demonstrates how attackers exploit these wireless deceptive techniques to compromise wireless liveness detection systems. Malicious attackers are capable of crafting fake wireless signals synchronized with spoofed video or audio streams to deceive systems into accepting fraudulent activities as legitimate. We refer to such attacks as phantom-CSI attacks. Experimental results on SDR platforms verify that these phantom-CSI attacks can significantly reduce the accuracy of spoof detection in wireless liveness systems.