RFNOID: Protecting RFID Motion Privacy via Metasurface

Yanni Yang, Zheng Shi, Zhenlin An, Runyu Pan, Yanling Bu, Guoming Zhang, Pengfei Hu, Jiannong Cao

Abstract: RFID tags have become prevailing human motion sensors in applications, like smart homes and health monitoring. However, RFID's powerful through-wall sensing capabilities raise significant privacy concerns about leaking human motion information, which the RFID sensing community has overlooked. To fill this gap, we propose RFNOID, the first system that can protect human motion privacy against adversarial through-wall RFID sensing. The key enabler is our design of an RFID metasurface consisting of multiple 1-bit phase shifters that can obfuscate motion information. As a pioneering work, we perform theoretical modeling and exploratory studies of the metasurface effect on RFID signals in temporal and spectral domains. Based on the preliminary analysis, we find it is non-trivial to achieve effective signal obfuscation in both domains due to the trade-off in simultaneously increasing temporal signal chaos and masking the human motion spectrum. To tackle this issue, we judiciously devise a metasurface controlling strategy by jointly optimizing the signal entropy, variance, and spectrum distribution to balance the temporal and spectral motion obfuscation. Extensive experiments demonstrate that RFNOID can significantly decrease the adversarial through-wall motion detection rate to less than 6% and increase the respiration estimation error by over 3×.

External IDs:dblp:conf/infocom/YangSAPBZ0025