Acoustic Eavesdropping from Passive Vibrations via mmWave Signals
Wei-Han Chen, Kannan Srinivasan · GLOBECOM 2022 - 2022 IEEE Global Communications Conference · 2022
Recovering audio signals from a vibrating source by analyzing fluctuations in the received wireless signals has been shown possible. However, previous RF-based works are limited to recovering sounds from an active sound source. In this paper, we present mmMic and demonstrate the feasibility of utilizing millimeter-wave (mmWave) signals to recover intelligible speech from a passively vibrating object. Such advance has a direct security implication of RF-based acoustic eavesdropping. We implement mmMic on a commercial-off-the-shelf (COTS) mmWave radio and validate its feasibility through real-world experiments. The results show that the objective quality of the recovered speech by mmMic is comparable with state-of-the-art systems.