| Summary: | 碩士 === 國立東華大學 === 資訊管理碩士學位學程 === 104 === With the comprehensive evolution of information communication technologies on mobile sensing objects, versatile ubiquitous networks embedded with specific-purpose sensors and intelligent wearable devices have promptly been identified, developed and deployed, called the Internet of Things (hereinafter, it is referred to as IoT). The advantages of data retrieval of modern intelligent objects, i.e. contactlessness and efficiency, bring a new era of IoT based application development. Meanwhile, the security issues have been mainly focused due to potential novel threats from IoT architectures. Recently, both the academic and industry have deeply investigated the design of transparent authentication on multi-modal networks. In consideration of the heterogeneous network property of IoT, in this thesis we propose an authentication system which applies machine learning techniques to extract user bio-features as authentication tokens and thus transparently performs continual or real-time entity verification in the background without the user’s notices.
|
|---|
