Cyber-Physical Systems (CPS) is the amalgamation of highly sophisticated sensors with physical spaces. These close conjunctions of sensors with communication infrastructure intrinsically linking to society’s Critical Infrastructures (C.I.) are being witnessed more often in the context of Smart Grid (SG). As a backbone of C.I., Smart Grid demonstrates ability to precisely monitor large scale energy systems and designed in order to achieve complex local and global objectives. Being capable of performing such sophisticated operation it also bears the vulnerability of being exposed for cyber-physical co-ordinated attack that may lead to catastrophic effect. Many researchers have analyze the different stages of cyber-physical co-ordinated attacks like attack detection, prevention, impact analysis and recovery plans but there exist a research gap to address all the issues under single framework. Through this paper, we propose a novel Cyber Physical Defense Framework (CPDF) based on National Institute of Standards and Technology (NIST) guidelines to address the cyber attack on SG. Our work addresses the pre and post attack scenario, attack vector formulation through hierarchical PetriNet modeling and recovery mechanism. We have performed experiment for Distributed Denial of Service (DDoS) and False Data Injection attack (FDI) to validate our framework effectiveness and established the efficacy of proposed model. In the end, we have presented a case study of FDI attack detection using machine learning technique on IEEE 9-bus and 14-bus system.
A Dynamic Strategy for Cyber-Attack Detection in Large-scale Power Systems via Output Clustering