CARACAS: Vehicular ArchitectuRe for Detailed Can Attacks Simulation

Modern vehicles are increasingly vulnerable to attacks that exploit networkinfrastructures, particularly the Controller Area Network (CAN) networks. Toeffectively counter such threats using contemporary tools like IntrusionDetection Systems (IDSs) based on data analysis and classification, largedatasets of CAN messages become imperative. This paper delves into thefeasibility of generating synthetic datasets by harnessing the modelingcapabilities of simulation frameworks such as Simulink coupled with a robustrepresentation of attack models to present CARACAS, a vehicular model,including component control via CAN messages and attack injection capabilities.CARACAS showcases the efficacy of this methodology, including a BatteryElectric Vehicle (BEV) model, and focuses on attacks targeting torque controlin two distinct scenarios.