The rapid evolution towards the Industry 4.0 improves the performances of Industrial Control Systems (ICSs). However, due to the unmanageable re-engineering cost of pre-existing industrial devices, insecure protocols continue to be used to manage these systems. In this scenario, legacy protocols, such as the Modbus/TCP, are still largely used to control a range of industrial processes alongside with modern technologies. Consequently, hybrid industrial infrastructures with both legacy and innovative devices require novel security and prevention methodologies.
In this work, we present AMON (Automaton MONitor): an Intrusion Detection System (IDS) based on Deterministic Finite Automata (DFA) for Modbus/TCP traffic monitoring. AMON combines DFA with the Longest Repeating Subsequence (LRS) algorithm, commonly used in bioinformatics, to model the traffic and identify anomalies. In order to address the challenges presented in hybrid scenarios, we extend AMON to work with the Constrained Application Protocol (CoAP), used for the Industrial Internet of Things (IIoT). We show preliminary results in a simulated industrial network and discuss possible implementation of the developed detection system to secure hybrid industrial infrastructures.
Conference: Proceedings of the 14th International Conference on Availability, Reliability and Security - ARES '19.
Date of Publication: August 2019.
Publisher: ACM Digital Library.