A Model-Driven Security Analysis Approach for 5G Communications in Industrial Systems

5G communication network has become a major pillar in the evolution of interconnected industrial systems. However, the introduction of 5G network may lead to unknown risks in the systems. To reveal the impact of network threats on 5G-based industrial systems, a 5G network security analysis approach combining formal modeling and attack penetration is proposed. Firstly, the 5G network models based on topology and transmission events are established to cope with diverse and hidden attack routes and behaviors. Then, the attack module is integrated into the network model. With attack penetration to the models, potential vulnerabilities are exploited and quantified based on the hierarchical-topology model, and network reliability is evaluated based on the transmission-event model. The simulation results identify and quantify network vulnerabilities under various attacks, including access authentication failure, destruction of data integrity, illegal control of Network Functions (NFs), and malicious consumption of shared slicing resources. Meanwhile, a more unpredictable outcome is that there is a threshold of access probability, a , to measure the impacts of attacks against the bearer network and core network on reliability. Finally, a practical case about the impact of network security on a 5G-based coupled-tank system is discussed, which further proves the feasibility of our approach.

Automated summary

This paper proposes a 5G network security analysis approach combining formal modeling and attack penetration to reveal the impact of network threats on 5G-based industrial systems. The approach establishes 5G network models based on topology and transmission events to handle diverse and hidden attack routes and behaviors. It integrates an attack module into the network model to exploit and quantify potential vulnerabilities. Simulation results identify and quantify network vulnerabilities under various attacks, and a threshold of access probability is found to measure the impacts of attacks on network reliability. The feasibility of the approach is further proven through a practical case study.