Mobile terminal user authentication scheme based on dynamic gesture

In order to guarantee the information security of the mobile terminal, a user authentication scheme based on a mobile terminal with a linear accelerometer embedded is put forward. It utilized an accelerometer to capture the acceleration signal of a person´s authenticating gesture. In the preprocessing part, a sliding-window and bottom-up(SWAB) algorithm was used to detect the valid gesture endpoint, and wavelet packet decomposition(WPA) method was used to filter the sensor power noises and random dithering. Then in the feature extraction part, the short-time energy feature which can fully characterize individual differences was chosen. Finally, an algorithm based on Gaussian Absolute Dynamic Time Warping(GADTW) kernel and One-Class Support Vector Machine(OCSVM) was proposed for authentication. The experimental results demonstrate that, in the case of imposter attack, the False Rejection Rate is below 6 percent while the False Acceptance Rate is approaching 0 percent, which show the effectiveness of the scheme.