Cyber-Physical System Defense Against Structured False Data Injection Attacks Using an Adaptive Security Framework with Passivity Enhancement

Abstract

System integrity, operation, and significant breakdowns can be compromised by coordinated False Data Injection Attacks (FDIAs), which are increasingly prevalent in Cyber-Physical Systems (CPS). Because they are dynamic and constantly evolving, these threats often bypass traditional security controls. The prompt identification of complex FDIAs, the reduction of anomaly detection false positives, and the maintenance of system stability in hostile environments are some important problems tackled. The Passivity-Enhanced Adaptive Security Framework (PEASF) is introduced in this work as a mechanism to enhance CPS security. PEASF integrates passivity-based control with adaptive security approaches to detect and neutralize real-time attacks. PEASF is engineered to suppress structured FDIAs by integrating passivity-based stability enforcement, adaptive intrusion detection, quantified attack impact analysis, and resilient control adaption. The framework employs hybrid detection methods to identify and measure attacks' effects reliably. These methods integrate graph models, machine learning classifiers, and Kalman filtering. Simulation analysis on a testbed of a CPS is conducted to evaluate the proposed PEASF framework concerning resilience against coordinated attacks, detection accuracy, and control adaptation efficiency. Relative to conventional control-based defense techniques, PEASF significantly enhances system stability, reduces detection errors, and enhances security resilience. The outcomes show that vital infrastructure fields like smart grids, intelligent transportation systems, and industrial automation can effectively apply PEASF to secure important power system components.