Continuous user authentication on smartphone via behavioral biometrics: a survey

Continuous user authentication on smartphone via behavioral biometrics: a survey

연구 분야: Analysis

학회: Multimedia Tools and Applications

The use of ubiquitous devices is increasing worldwide due to signs of progress in sophisticated hardware technology with cutting-edge features. These features offer most users to store their personal information on the device. One of the most usable and off-the-shelf ubiquitous devices in our daily lives is the smartphone. To protect the stored information, the smartphone is enabled with entry-point authentication. Moreover, existing methods of entry-point authentication are not useful unless access control is vigilantly protected. To overcome these issues, the current trend in user authentication of the smartphone is implicit or continuous authentication. In the coming years, the smartphone is expected to enable with intelligent access control in support of continuous authentication. This article reviews available public datasets in continuous authentication of the smartphone. Then, we delve into unimodal and multimodal aspects of continuous authentication. We review the available literature on both aspects of continuous authentication that describe problems tackled, the methodology used, associated datasets, and approaches used for performance evaluation. At the end of each aspect, we highlight the lessons learned based on the literature. Furthermore, this article also reviews attacks on behavioral biometrics of the smartphone. Finally, this article confers open challenges and potential directions for future exploration that must be addressed. This article helps readers to understand the datasets, existing methods in this field, and fosters to implement new methods without pitfalls.