• attacker identification
• automotive
• cybersecurity
• ECU fingerprinting

Cars are getting smarter, so much that they have become “connected computers on wheels”. The brain of a vehicle is made up of electronic control units (ECUs) that communicate with each other, exchanging information and data and thus managing all the main components. The control units, moreover, can also be connected with external devices through interfaces such as Wi-fi, USB, Bluetooth etc. which increase the surfaces vulnerable to potential attacks. Various defence systems that determine the presence of an attack on the in-vehicle network have been proposed. They, however, fail to identify which ECU transmits the attacking message, an essential point to isolate and counter an attack. This is the reason why fingerprinting comes into play: a technique that characterizes the ECUs, making them uniquely identifiable. Therefore, this thesis aims to analyse, design and implement an innovative method that identifies if there is a cyber attack on in-vehicle networks and which is the attacking ECU. To this aim physical (voltage) measurements on the in-vehicle network (e.g. CAN) are used combined with advanced digital processing algorithms to make cars more secure.