Abstract
While the smart phones become the one of the most popular technological products, the security issues of the smart phone are more and more important. One security issue about the smart phones is the screen lock. Nowadays the screen locks on Android which are widely used are the PIN codes and the graphical screen lock called Android Pattern Lock based on PassShapes. The PIN codes are combination of the numbers; Android Pattern Lock is 9 points that are arranged as the 3 by 3 grid. A user employs his fingers to draw a shape on these 9 points. Android Pattern Lock is easy to memorize than the PIN codes. However, Android Pattern Lock cannot resist Smudge Attacks and Shoulder-Surfing Attacks. And then, the complexity of Android Pattern Lock is lower than the PIN codes. A new screen lock called ShakePass is proposed in the thesis. A user can shake his smart phone to draw a shape as his graphical password. The approach utilizes the accelerometer (G-Sensor) on the smart phones to detect the directions of the user shakes. Because this approach is not limited by 9 points, the complexity of ShakePass is higher than Android Pattern Lock. Moreover, ShakePass uses the concept of VibraPass to confront Shoulder-Surfing Attacks. When a user starts to input his password, ShakePass can randomly turn on the long vibration to prompt the user to shake a fake direction. Therefore, the attackers do not know which direction is true or false. The new screen lock approach, ShakePass, can solve the problems as mentioned above, and increase the complexity.