Continuous user authentication in mobile device and IoT environment

• Main Author: Liu, Allen Yang
• Other Authors: Ng Wee Keong
• Format: Thesis-Master by Research
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Computer science and engineering::Information systems::Information systems applications
• Online Access: https://hdl.handle.net/10356/143089
• Institution: Nanyang Technological University

The conventional authentication methods, such as password authentication, fingerprint recognition and face recognition, have many drawbacks and disadvantages. The conventional authentication methods are snapshot and one-time authentication with a very high risk of being hacked. In the early days, the hackers can cause $2,300 in fraud losses every hour by phishing or other hacking methods. The password leak and account takeover are serious security problems in financial industry. We aim to create continuous authentication methods in mobile devices and IoT environment which allows user to authenticate without the burden of credential requests. Thus, creating a secure, user-friendly experience for the user. We have proposed two methods to authenticate users in two different scenarios, which are mobile device environment and IoT environment. Both of the two methods are proposed to authenticate user in the background by learning their behaviours, and building their unique profiles. And machine learning techniques are applied for information discovery from the behaviour data to make authentication decisions.