A Hybrid Approach to Cyber-Physical System Security Using AI-Based Intrusion Detection and Blockchain

Abstract

The core of critical infrastructures is represented by Cyber-Physical Systems (CPS), but with its close coupling of the real-world to networked control, they are sought after by advanced cyber-attacks. Current intrusion detection systems and data integrity solutions tend to work independently and are not able to offer intelligence and trust. In this paper, it is suggested to develop a hybrid security system that combines the Artificial Intelligence (AI)-based intrusion detection and the Blockchain-used data protection to enhance CPS security. Anomalous behavior is detected by the AI module, and the blockchain layer is used to guarantee immutable logging and decentralized access control. Simulated smart-grid CPS experimental assessment proves greater detection accuracy, fewer false positives, and greater resistance to data corruption when compared with traditional methods.