G-Key: An Authentication Technique for Mobile Devices Based on Gravity Sensors
Linjiang Xie, Hequn Xian, Xuyue Tang, Wei Min Guo, Feilu Hang, Ning Fang · 2019
User authentication is an essential issue of information security and privacy protection in mobile device based applications. Traditional password authentication causes a heavy memory burden. Biometric authentication on mobile devices are confronted with privacy issues and risks of information disclosure. Based on the characteristics of the built-in sensors of mobile devices, we propose G-key, a novel authentication method. Primitive postures are defined, which can be identified by reading of gravity sensors of the device. A G-key recognition algorithm is designed, in which parameters can be adjusted to achieve balance between accuracy and efficiency. Users can arbitrarily define a private gesture key from posture combinations as their authentication basis. Compared with biometric authentication and password-based authentication, the authentication process of G-Key is more private. It does not require storing sensitive biological information of individuals, and no memory burden is involved. We implement G-Key in the Android platform, and experiments show that it works with high accuracy.