Touch Biometrics for Unobtrusive Continuous Authentication on Smartphones

Abstract

Mobile devices such as smartphones have until now been protected by traditional authentication methods, including passwords or PINs. These authentication mechanisms are di cult to remember and often disabled, leaving the device vulnerable if stolen. This dissertation presents an unobtrusive, continuous authentication system based on biometric data collected from a touchscreen. A model is presented for implementation of such a touch biometric system on a smartphone. This model is evaluated by conducting an experiment simulating real-world touch interaction. Touch data was collected from 30 participants during normal phone use. This data was analysed using two classi cation algorithms to determine the accuracy of matching touch data to users. The results show that touch data is distinct between users and could be used in authentication without disrupting normal touch interaction. Furthermore, touch data could be used in its raw form without signi cant pre-processing.