A Behavioral Modeling-based Driver Authentication Approach for Smart Cars Self-Surveillance

Driver authentication is a vital aspect that enhances passengers' safety and security. Indeed, driver authentication enables detection of maintaining normal driving behavior, car theft, fraudulent switching of designated drivers prevention, etc., which allows a vehicle system to distinguish between legitimate users from not legitimate ones and make transportation safer. This paper introduces an efficient in-car driver authentication approach for vehicle security and safety based on driver behavior. The proposed approach continuously monitors the driver's behavior and compares it with normal behaviors already recorded during a learning phase. Our solution is based on an algebraic method, namely the Conditional Basic Process Algebra (CBPA), to model the driver's driving style in two stages, before and after starting the car, taking into account the positions of the seat, mirrors, the seat belt, etc. We demonstrate through simulation results the efficiency of our approach in terms of response time and detection success rate.