Over recent years, smart locks have evolved as cyber-physical devices that can be operated by digital keypads, physiological biometrics sensors, smart-card readers, or mobile devices pairing, to secure door access. However, the underlying authentication schemes, i.e., knowledge-based (e.g., PIN/passwords), possession-based (e.g., smartphones, smart cards), or physiological biometric-based (e.g., fingerprint, face), utilized in smart locks, have shown several drawbacks. Studies have determined that these authentication schemes are vulnerable to various attacks as well as lack usability. This paper presents SmartHandle - a novel behavioral biometric-based transparent user authentication scheme for smart locks that exploits users’ hand-movement while they rotate the door handle to unlock the door. More specifically, our solution models the user’s hand-movement in 3-dimensional space by fetching the X, Y, and Z coordinates from 3 sensors, namely, accelerometer, magnetometer, and gyroscope corresponding to the hand-movement trajectory, to generate a user-identification signature. We validated our solution for a multi-class classification scenario and achieve a True Acceptance Rate (TAR) of 87.27% at the False Acceptance Rate (FAR) of 1.39% with the Linear Discriminant Classifier (LDC) on our collected dataset from 11 users. The solution can be easily deployed at the main entrance of homes and offices offering a secure and usable authentication scheme to their legitimate users.
SmartHandle: A Novel Behavioral Biometric-based Authentication Scheme for Smart Lock Systems
The work described in this website has been conducted within the project NeCS. This project has received funding from the European Union’s Horizon 2020 (H2020) research and innovation programme under the Grant Agreement no 675320. This website and the content displayed in it do not represent the opinion of the European Union, and the European Union is not responsible for any use that might be made of its content.