Project Title: Identification and Mitigation of Coordinated Attacks on Distributed Energy Management
Project Summary: Electric energy delivery systems (EDS) of the future will contain millions of intelligent embedded sensors and control devices that will allow the system to operate in a distributed and decentralized manner. These two concepts: distributed and decentralized enable far more resilient operation and control while supporting the integration of loads and renewable generation in a plug-and-play manner. Achieving these goals require parallel developments in communication, estimation and control which are actively being developed. It is critical to realize that to distributed intelligence and decentralization requires coordination and cooperation between devices to ensure that both local and global operational constraints are met. Although it might be easy to design schemes to ensure graceful performance degradation for the failure of one or ore devices, the securing of the distributed EDS is very challenging because: (1) EDS comprise a large number of cyber and physical components that exhibit highly complex inter-dependencies, (2) it may be prohibitively expensive to secure all sensors, (3) encryption/decryption schemes introduce latencies that may not be acceptable for all control functions, and (4) many field devices operate in an unsupervised manner and may have limited computational resources because of the distributed nature.
The project will develop a formal framework to characterize the attack space by verifying the feasibility of an attack, determining its consequences, identifying adversarial attributes, and determining how these are influenced by inter-dependencies between the cyber/computing and physical components. The proposed framework also determines the most cost-effective risk mitigation considering the security benefit. The research will also explore novel smart grid agility techniques to counter “tactical” dual-stage attacks, during which an adversary first probes the system to learn about the resilience model, and then launches a potent evasive attack. Additionally, the proposed research addresses the resilience of EDS and automatic synthesis of counter-measures using deception to thwart attacks under uncertainty by obfuscating and shifting the attack surface.
University Team Members
Lead: Dr. Badrul Chowdhury – UNC Charlotte
Aranya Chakrabortty – NC State University
Ehab Al-Shaer – CYberDNA Security
Industry Advisor: David Lawrence – Duke Energy