An analytical survey of cyber‐physical systems in water treatment and distribution: Security challenges, intrusion detection, and future directions

Abstract

AbstractSince the inception of the Industrial 4.0 revolution, industrial cyber‐physical systems (CPSs) have become integral to critical infrastructures and industrial sectors, including water treatment and distribution systems. Integrating physical and digital worlds has made communication systems within these plants—comprising actuators, sensors, and controllers—vulnerable to advanced cyber‐attacks. Safeguarding the nation's critical infrastructure has thus attracted significant interest from both academia and industry. This article thoroughly examines water treatment and distribution CPSs, detailing their architectural design, devices, applications, and security standards. It analyzes various cyber‐attacks and explores CPS security vulnerabilities and their detection and mitigation techniques. Additionally, it reviews the trends in machine learning (ML) and deep learning (DL) intrusion detection system (IDS) solutions, highlighting their advantages and disadvantages. The article evaluates current datasets and testbeds, identifying some of the best‐performing IDS algorithms tested on each dataset compared to previous research, which could serve as benchmarks in this field. Finally, it proposes data augmentation techniques to generate comprehensive datasets, identifies research gaps, and suggests potential improvements to enhance IDS performance.