DCUS : evaluating double-click-based unlocking scheme on smartphones

Abstract

With the increasing capability of software and hardware, mobile devices especially smartphones are changing the way of peoples’ communication and living styles. For the sake of convenience, people often store a lot of personal data like images on the device and use it for completing sensitive tasks like payment and financial transfer. This makes data protection more important on smartphones. To secure the device from unauthorized access, one simple and efficient method is to design a device or screen unlock mechanism, which can authenticate the identity of current user. However, most existing unlock schemes can be compromised if an attacker gets the correct pattern. In this work, we advocate that behavioral biometrics can be useful to improve the security of unlock mechanisms. We thus design DCUS, a double-click-based unlocking scheme on smartphones, which requires users to unlock the device by double clicking on the right location on an image. For user authentication, our scheme needs to check the selected images, image location and double-click patterns. In the evaluation, we perform a user study with 60 participants and make a comparison between our scheme and a similar unlock scheme. With several typical supervised classifiers, it is found that participants can perform well under our scheme.