Title: On fuzzy inference based supervisory control decision model with quantum artificial intelligence electromagnetic prediction models

Authors: Varghese Mathew Vaidyan; Akhilesh Tyagi

Addresses: Department of Electrical and Computer Engineering, Iowa State University, Ames, IA, USA ' Department of Electrical and Computer Engineering, Iowa State University, Ames, IA, USA

Abstract: Robust Supervisory Control Systems that interface with programmable logic controllers (PLCs) are crucial components of Industrial Control Systems (ICS) to manage critical infrastructure. On the other hand, current industrial control systems consist of highly connected Internet of Things (IoT) nodes. This makes them a target of malware written in ladder logic or another IEC 61131-3-compliant language. Due to their networked nature, Supervisory Control systems are potentially subject to widespread cyberattacks. To manipulate the Supervisory control function, a denial of service attack on a single Real Time Unit (RTU) might be leveraged. The vulnerability of Supervisory control systems is a well-known concern in industrial control. Thus, in the case of a coordinated attack on IoT control systems, the dependability of Supervisory control decision-making processes is jeopardised. We propose a decoupled fuzzy inference based Supervisory Control Decision Model which is based on electromagnetic (EM) emissions from real-time units. A hybrid quantum artificial intelligence model derived from EM emissions adds another layer of reliability to traditional Supervisory control systems, in the event of widespread IoT adversarial attacks. Our objective is to minimise industrial plant downtime and false alarms through Fuzzy Inference-based localisation of adversarial attacks/faults, risk severity levels, and impacted nodes via EM based Hybrid Quantum AI (QAI) fault analysis. We investigated fuzzy topologies, attack routes, and faults and found that our Fuzzy Inference based Supervisory decision-making approach is capable of operating independently of the traditional Supervisory Control network, localising the attack, and therefore minimising downtime and costs.

Keywords: fuzzy inference systems; supervisory control; deep learning; cyber-physical systems; industry applications; PLC; logic bombs; quantum computing; electromagnetics; fault analysis; control systems.

DOI: 10.1504/IJCCPS.2023.133732

International Journal of Cybernetics and Cyber-Physical Systems, 2023 Vol.1 No.3, pp.261 - 276

Received: 06 May 2022
Accepted: 23 Aug 2022
Published online: 02 Oct 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article