PressHeart: A Two-Factor Authentication Mechanism via PPG Signals for Wearable Devices

The integration of biometric-based user authentication into wearable devices has become increasingly important for protecting users' private information and property. In this paper, we propose a two-factor authentication mechanism, PressHeart, which utilizes widely-used Photoplethysmography (PPG) sensors embedded in wearable devices. Our observations reveal that PPG sensors can implicitly measure excitation press signals when the users press the skin or the device, which implies individual wearing and behavioral habits that can serve as reliable factors for user authentication. For better separating the press signals from PPG signals and extracting sufficient signals for user authentication, we introduce two adaptive segmentation methods and a specific feature set for feature extraction in PressHeart. To validate the performance of Press Heart, we develop a prototype with a PPG sensor and conduct experiments involving 14 participants. The experiment results demonstrate that PressHeart can achieve an average of 94.9 % accuracy with high authentication efficiency and security.