Threat Modeling to Secure <scp>CPS</scp> ‐Based Critical Infrastructures

ABSTRACT Cyber‐physical systems (CPS) are critical components of many vital infrastructures, including hydro power plants (HPPs). However, these systems are susceptible to various cyberattacks, as evidenced by past attacks. This research paper explores CPS security in HPP environments, presenting a review of existing architectures and threat modeling techniques. Considering the importance of the availability of CPS‐based critical infrastructures, we have proposed a redundant HPP‐based architecture containing diverse vendor components to improve system resilience and security against both common and zero‐day vulnerabilities. Additionally, a novel CritFit Threat Modeling Framework (CritFit‐TMF) is proposed by integrating STRIDE, DREAD, and CritFit Scoring. The STRIDE and DREAD threat modeling approaches are intended to identify and evaluate the CPS threats, respectively, whereas the CritFit Scoring is applied to the outcome of the DREAD model so as to contextualize the priorities of security aspects with respect to CPS. The CritFit Scoring methodology is to tailor severity assessments specifically for CPS, recognizing the distinct security priorities of CPS compared to traditional IT systems. This proposed CritFit‐TMF enhances the identification, evaluation, and prioritization of threats during the design phase of HPP‐based CPS and thus reduces the cost and complexity of patching vulnerabilities during the development or deployment phase of CPS. The proposed CritFit‐TMF is simulated by considering our proposed redundant HPP architecture. The evaluations highlight the feasibility, adaptability, and importance of our proposed framework in CPS. Simulation results of the CritFit‐TMF revealed that 24.90% of the identified threats were Elevation of Privilege, followed by Denial of Service (24.08%), Spoofing (23.27%), Information Disclosure (10.2%), Repudiation (9.39%), and Tampering (8.16%).