Synchronous Fusion for Continuous Biometrics Authentication

Existing authentication techniques are vulnerable to many attacks and abuses, such as password-capturing spyware, network sniffing, social engineering and unauthorized sharing. In this context, a new identity verification technology termed continuous authentication uses behavioural biometrics to ensure that the user is not using an identity that has been highjacked or stolen.This is performed by monitoring and processing periodically biometrics infromation from human computer interactions data such as mouse movements and keystrokes. Mouse and keystroke biometrics scores are produced asynchronously during the continuous process, making it hard for multimodal system involving both modalities to take full advantage of their combined strengths. The objective of the current MITACS proposal is to research and develop a dynamic and adaptive biometric fusion technique that will address the above limitation by allowing effective combination of mouse and keystroke biometrics schemes during continuous authentication. The new fusion model will be implemented in a future release of a continuous authentication product named Bio Tracker.

Faculty Supervisor:

Dr. Mihai Sima

Student:

Youssef Nakkabi

Partner:

Plurilock Security Solutions Inc.

Discipline:

Engineering

Sector:

Information and communications technologies

University:

University of Victoria