Audio Hotspot Attack: An Attack on Voice Assistance Systems Using Directional Sound Beams and its Feasibility
Ryo Iijima,Shota Minami,Yunao Zhou,Tatsuya Takehisa,Takeshi Takahashi,Yasuhiro Oikawa,Tatsuya Mori
DOI: https://doi.org/10.1109/tetc.2019.2953041
2021-10-01
IEEE Transactions on Emerging Topics in Computing
Abstract:We propose a novel attack, called an "Audio Hotspot Attack," which performs an inaudible malicious voice command attack, by targeting voice assistance systems, e.g., smart speakers or in-car navigation systems. The key idea of the approach is to leverage directional sound beams generated from parametric loudspeakers, which emit amplitude-modulated ultrasounds that will be self-demodulated in the air. Our work goes beyond the previous studies of inaudible voice command attack in the following three aspects: (1) the attack can succeed on a long distance (3.5 meters in a small room, and 12 meters in a long hallway), (2) it can control the spot of the audible area by using two directional sound beams, which consist of a carrier wave and a sideband wave, and (3) the proposed attack leverages a physical phenomenon i.e., non-linearity in the air, to attack voice assistance systems. To evaluate the feasibility of the attack, we performed extensive in-lab experiments and a user study involving 20 participants. The results demonstrated that the attack was feasible in a real-world setting. We discussed the extent of the threat, as well as the possible countermeasures against the attack.
computer science, information systems,telecommunications