Automatic

Construction of Model-based Diagnosis of Cyber-Attack in Cyber-Physical Systems Using Labeled Petri Nets

Publié le - 2023 IEEE 28th International Conference on Emerging Technologies and Factory Automation (ETFA)

Auteurs : Romain Thibert, Gregory Faraut, Saïd Amari

Cyber-Physical Systems (CPS) brought connectivity to factories, and with connectivity comes a risk of cyber attack. CPSs are vulnerable to malicious attacks in which an attacker is inserted between a process and its control unit. Some papers have proposed to model attacked CPSs with discrete event systems (DES), have successfully modeled attacked systems and characterised different types of attacks. Nevertheless, these papers tend to treat attacks after they have become problematic. The objective of this paper is to extend previous works on cyber attack in DES to build models for diagnosis of attacks. We will design a model using Labeled Petri Nets and construct a reachability graph of an attacked net in order to enrich a diagnosis model which will allow the detection of attacks before they completely destabilise the system. This construction is illustrated by an industrial example.