Security Risk Assessment for Connected Vehicles Based on Back Propagation Neural Network

The wide application of information communication technology makes connected vehicles (CV) more vulnerable to be attacked. The communication systems and key nodes of CVs, e.g., electronic control units, controller area network (CAN), in-vehicle infotainment system, will face various threats such as eavesdropping, tampering, and counterfeiting. Cyber security of vehicles should be paid more attention because it involves drivers' safety and public traffic security. It is impossible to deal with all kinds of security threats for the effect of performance, compatibility, cost, and efficiency of vehicles. Therefore, it is necessary to assess the security risk of CVs. This paper proposes a security risk assessment method based on the conventional security risk analysis model and utilized back propagation (BP) neural network. The simulation results show that the risk level of CVs can be evaluated quantitatively by trained neural networks, and the method is convenience and applicability for security risk assessment of vehicles.