False Data Injection in Scalar Feedback systems
Department: Electrical and Computer Science
Advisor: Parv Venkitasubramaniam
View: UGRS Research Poster (PDF)
Feedback systems exist throughout the modern world, controlling everything from cars, planes and shipping to electronic systems and power plants. These feedback systems are increasingly connected to the Internet for data collection and optimization purposes. This connectivity increases risk of adversaries attacking the system remotely. A potential security weakness in a standard feedback system is the control unit. Should an adversary, such as an Internet hacker, begin dictating control policy, they could have catastrophic effects on the system. Our research involved MATLAB simulation of the effect of attack policies on the Linear Quadratic Gaussian feedback model to determine the most effective methods. These simulations determined that constant attack policies do not yield high cost, and that the most effective attack policy was a white Gaussian noise policy.
About Eric Reid:
Eric Reid is junior studying electrical engineering at Lehigh University, and hopes to pursue his PhD in electrical engineering after graduating. Eric is interested in circuit design, fabrication and associated fields. He is a project manager of the student IEEE chapter at Lehigh university, and is a member of Eta Kappa Nu and Kappa Kappa Psi National Honorary Band Fraternity. Eric began doing simulation with Matlab simulation in the June of 2015.