Mississippi State University
Jones, Bryan A.
Morris, Thomas H.
Date of Degree
Graduate Thesis - Open Access
Master of Science
James Worth Bagley College of Engineering
Department of Electrical and Computer Engineering
Supervisory control and data acquisition systems are extensively used in the critical infrastructure domain for controlling and managing large-scale industrial applications. This thesis presents a security management structure developed to protect ICS networks from security intrusions. This structure is formed by a combination of several modules for monitoring system-utilization parameters, data processing, detection of known attacks, forensic analysis to support against unknown attacks, estimation of control system-specific variables, and launch of appropriate protection methods. The best protection method to launch in case of an attack is chosen by a multi-criteria analysis controller based on operational costs and efficiency. A time-series ARIMA model is utilized to estimate the future state of the system and to protect it against cyber intrusions. Signature and performance based detection techniques assist in real-time identification of attacks with little or no human intervention. Simulation results for Scanning, Denial of Service and Injection attacks are provided.
Trivedi, Madhulika, "Toward Autonomic Security for Industrial Control Systems" (2012). Theses and Dissertations. 4751.