In modern vehicles, a large number of Electronic Control Units (ECUs) are responsible for monitoring and controlling the state of the car. ECUs communicate using the Controller Area Network (CAN) protocol and can interact with the external world. This has increased the attack surface of vehicles and created the opportunity for new attacks. The objective of the thesis is the design, development and validation of a software framework for a multi-features intrusion detection system (IDS) suitable for attacks detection in vehicles.
The multi-features IDS is compliant to the AUTOSAR standard, and integrates developed IDSs based on a rule based approach and ECU fingerprinting techniques, including voltage and timing techniques. The main functionalities of the framework are the intrusion detection system manager of anomalies, the filtering and storage on a security event memory. Being compliant with AUTOSAR, the developed multi-feature IDS can be easily integrated in vehicle embedded software. Test results demonstrate its effectiveness in reducing cyber-threats, contributing to the growth of the automotive cyber-security field.