Continuous User Authentication in Mobile Phone Browser Based on Gesture Characteristics

Conventional authentication methods like passwords, PINs, graphical keys do not provide the required level of security. Those authentication methods might be defeated by peeping or using keyloggers to reveal PINs and passwords. Using authentication by fingerprint also could be vulnerable in case an intruder seizes mobile phone after an authentication procedure. Due to that weakness of static mobile user authenti-cation approaches the continuous methods were researched. The approach to mobile phone user authentication via continuous gesture recognition is proposed in this article. During the research trust level evaluation model for continuous user authentication was developed and a range of novelty detection algorithms using parameters of user gesture were tested. As a result, a security module for a mobile phone browser was created.