Cognitive Authentication Mechanisms for User Verification in Autonomous Vehicles

Abstract

In the case of the smart attack, it is shown that for even highly secure DAF designs, drivers can be impersonated through emulation and cloning. Another research gap highlighted is the lack of standard methodologies for testing and evaluating behavior-based defense systems, especially in the context of vehicular architectures. Our paper demonstrates as a pilot work, that the proposed methodology can be synthesic, in the sense of proof-of-concept case studies and validate through static and dynamic security evaluations how the use adversarial simulations can systematically disclose the level of security in such locations. In the future, our research could be verified and prolonged by using this testing framework on the border environment with users having cognitive impairment or/and voice issues.