Adopting Behavioral Characteristics in Pattern Unlocker

Abstract

Security of smartphone has been concerned by many users, because lots of personal data are saved locally in the phone. Although traditional screen locks have introduced, offenders can still crack in the phone by observing the visible oil marks on the screen to unlock it. The safety measures seem useless. To improve the security level of screen lock, the project aims to develop an enhanced version of pattern screen lock in Android platform, which requires verification of user’s behavioural characteristics. The application stores the spatial coordinates and press pressure when user input the pattern. By using Dynamic Time Warping algorithm, similarity of different patterns can be measured. We can then identify and authenticate a user by verifying these behavioural characteristics in his/her patterns.