Alessio earned a Master’s degree in Computer Engineering in October 2018 at Università degli Studi Roma Tre. I joined Université du Luxembourg and SP2 as a PhD candidate on January 15, 2019.

Alessio’s main research interests are cybersecurity, privacy and Machine Learning. His current research focuses on cybersecurity for intra-vehicle communication, which is critical for the safety of the driver and other agents on the road.

CAN Bus is by far the most adopted protocol for intra-vehicle communication. The Electronic Control Units (ECUs) are the devices connected to the bus, which perform different tasks (e.g. steering wheel, brake pedal etc.). CAN bus works as a peer-to-peer network. CAN Bus was designed without any authentication protocol nor encryption. For this reason, any attacker with physical access to the vehicle can plug his own device and send forged data.

In response to this menace, introducing authentication based on cryptography seems not feasible, due to the low bandwidth of the CAN Bus. On the contrary, applying Machine Learning algorithms for intrusion detection and prevention might be an efficient way to provide cybersecurity. A classifier is trained to recognize/fingerprint local ECUs through physical characteristics of the ECUs. After a model is created and tested, it can be used for intrusion detection.

Until now no intrusion detection system has achieved a perfect accuracy respecting all the needed requirements. Alessio’s future work consists in implementing a new and better intrusion detection system on CAN bus and other currently adopted protocols for in-vehicle communication, in order to enhance cybersecurity in the vehicles.