Analisi ibrida delle minacce = Hybrid Threat Analysis

Nowadays, cyber-physical systems are ubiquitous and present in most of our everyday life. Their usage is spread over many applications, like industrial control systems, modern buildings, and critical infrastructures. Recently such systems are becoming more interconnected and smart, by means of the Internet of Things (IoT) concept, making them accessible through both physical and remote connections. Furthermore, end users can configure and customize their own systems, but this flexibility generally implies an unaware sharing of private data over the underlying IoT. The attack surface is thus much wider and more heterogeneous. Determining the potential threats in these smart systems and to provide an adequate amount of protection becomes a very complex and hard process. Generally, threat analysis covers partially the attack surface, because it does not consider the possible relationships among physical access, remote connections and human interactions, which may lead to unknown vulnerabilities. The next phase of risk mitigation is conducted to lower the risk of the most highly impacted threats only. This approach, besides being cumbersome and time-consuming, lacks a comprehensive view of the system behavior. It results, as such, error-prone, and it can make the overall system security significantly hard to maintain. The objective of the thesis is to define and implement a framework able to analyze any connections (physical or remote) and human-system interactions that may lead to a vulnerability, which is a task that the existing approaches cannot do.