Detecting Compromised Devices

PI: A. Selcuk Uluagac, Florida International University

Compromised devices can cause serious damages to the energy system as demonstrated in Stuxnet. This project aims to detect compromised devices in the smart grid domain. The main idea of the project stems from the fact that the compromised devices’ system calls are different from good devices due the extra functionality of malware. Hence, this project will design and implement a system-level configurable software tool that is capable of monitoring and detecting compromised smart grid devices.

Specifically, the software will combine system call tracing (i.e., ptrace, library interposition) and statistical and machine learning techniques to monitor and detect compromised device behavior. The tool can be deployed into devices in the smart-grid through firmware update.