DEVELOPMENT OF A CONCEPT OF A DIGITAL TWIN FRAMEWORK FOR CYBER-PHYSICAL SYSTEM

Abstract

In this article, the concept of a digital twin framework for multi-component cyber-physical systems, which will allow creating a security model based on a stochastic Petri net, was described. This framework consists of five connected logical layers. The physical layer includes physical system along with all sensors and actuators that collect security-related data from the system and generate a digital twin to control it through the feedback generation module. The PT-DT layer is added for storing and managing both security-related and general data that comes from the physical environment and is used by the digital twin to create the simulation model. The Digital Twin layer encompasses four key components: System Modeling, Simulation, Feedback Generation, and Visualization. In security modeling, attack graphs, attack trees, and Petri Nets are commonly used for modeling adversarial behavior, considering real assets and vulnerabilities. The DT-SERV layer gathers and organizes the specific knowledge and data required to develop, customize, and optimize higher-level services. This includes system and security rules that can be used for instance for attack/intrusion/anomaly detection. The service layer includes main services that can be classified by exploited operational mode. DTs can be executed in different operational modes, i.e., simulation, analysis, and replication based on what kind of data they use. Thus, the implementation of the framework will enable the identification of vulnerabilities in system components and the development of countermeasures to address attacks in the physical environment. This approach will help in proactively securing the system by detecting security incidents, analyzing their impact, mitigating potential risks, and preparing appropriate responses to address emerging threats effectively.