Curriculum
Computer Science and Innovation for Societal Challenges, XXXV series
Grant sponsor
Fondazione CARIPARO and Yarix srl
Supervisor
s
Mauro Conti
Co-supervisor
s
Luciano Gamberini

Project: Cybersecurity of Modern Cyber-Physical Systems
Full text of the dissertation book can be downloaded from: https://www.research.unipd.it/handle/11577/3475902

Abstract: Cyber-Physical Systems (CPSs) refer to those systems characterized by the interconnection of information technology and the physical process domains. These systems are nowadays employed in a wide range of applications, such as health monitoring, industrial control systems, and transportation. The recent digitalization and smartification of the processes required to integrate the Internet connection into CPSs, enabling functions like remote connection and cloud computing, but at the same opening new dangerous vulnerabilities surfaces. Indeed, recent events in history have shown many cyber-attacks and vulnerabilities discovered on CPSs. For this reason, there is still a need to contribute to securing such systems, both from the design and implementation points of view. In this thesis, we analyze the cybersecurity of modern CPSs, identifying and highlighting the current vulnerabilities, the research gaps in terms of security, and the threats affecting them. Then, we propose novel security mechanisms to prevent potential cyber-attacks. This thesis is composed of three parts as follows. In the first part of the thesis, we will focus on the security of Industrial Control Systems (ICSs). These systems are used to control and monitor critical infrastructures and industrial processes. As a first step, we gather all the knowledge in this field from the literature, and we provide a systematic analysis of the testing platform and the detection systems solutions operating on them. To motivate the necessity of improving the security of current industrial systems, we performed a measurement study highlighting the dramatic exposure of the communication protocols and services of more than $50$ industrial endpoints. Then, we developed and deployed an innovative ICS honeypot. While measuring the honeypot exposure, we noted that industrial systems are still highly targeted and interacted with by malicious actors over the internet on specific vulnerable industrial services. In the second part of the thesis, we will look at the security of vehicular systems. Like ICSs, modern vehicles present vulnerabilities due to the adoption of legacy components, enabling the possibility of malicious exploits. To this end, we will focus on the internal communication bus of cars, we examine its vulnerabilities, the current solutions in the literature, and their limitations, and propose an innovative cryptographic key distribution system among the network nodes. We will then focus on the emerging electric vehicle paradigm. We identified two possible cyber-attacks on this ecosystem. The first is based on a relay attack vulnerability, which implies charging illegitimate vehicle recharging fees. Instead, the second one consists of a privacy leakage from the current absorbed during the vehicle's recharging process. In the third part of the thesis, we leverage the knowledge of our studies to investigate the security of CPS cross-domain applications. In particular, we first present a survey on Power Side-Channel (PSC) exploits in the literature, focusing on existing attacks and countermeasures. Indeed, PSCs have been proven effective in reversing and profiling the functioning of many embedded devices (e.g., smart cards, vehicles, and laptops). Then, we develop a novel framework to fingerprint Universal Serial Bus devices from their power consumption. This funding can be used, for instance, to securely authenticate a personal device and avoid malware delivery injection in critical applications (e.g., Stuxnet). Finally, we present the first security analysis of the emerging Hyperloop transportation technology. Hyperloop merges the concepts of ICS since it consists of a critical, distributed, and sensing infrastructure, and the concept of vehicle, due to the pod communication management. As a result, Hyperloop inherits all the vulnerabilities and risks of the two systems