Phase-Proof: Robust Mobile Two-Factor Authentication Via Phase Fingerprinting

The two-factor authentication (2FA) has been widely applied with the proliferation of mobile devices. Currently, many existing 2FA solutions propose to use acoustic fingerprints as the second factor. However, these schemes mainly consider using the magnitude or non-linearity information of the acoustic signal for authentication and ignore the fingerprint variations caused by the change of transmission distance between devices, which could threaten the accuracy of the system. To address these vulnerabilities, we propose Phase-Proof, a secure and robust 2FA that utilizes the phase distortions incurred by the acoustic elements of mobile devices as the second proof. Given the received acoustic signal, our system first extracts phase information from the signal and designs a new distance effect elimination scheme to remove the impact of transmission distances for robust fingerprint extraction. Moreover, our device authentication component then adopts a transfer learning-based method to capture the subtle differences in devices' fingerprints for accurate device authentication. Moreover, to further withstand the co-located attacks, our proximity detection component collects certain environmental randomness from the enrolled phone, and then matches it with the environmental randomness collected from the login device to verify the proximity of two devices. Our experimental results show that our system is accurate in providing 2FA and robust to various attacks across different scenarios.