In-depth cyber-physical attack detection via smart devices and deep learning

Project Supported by Public Organizations in Other Countries, 2021 - 2025

Project Type: Project Supported by Public Organizations in Other Countries
Begin Date: December 2021
End Date: January 2025

Project Abstract

The attack models we are considering for the proposed cybersecurity analysis and design involve the compromise of a set of smart devices on the network. Once a device is compromised, a sophisticated attacker may recover keying material, replicate the physical communication medium interface, and take over both the control and metadata signaling of the device. The proposed approach is to : 1) Monitor both cyber and physical data from smart meters, PMUs, smart PV inverters, and other existing smart devices in the network. 2) Conduct a proactive, predictive, and effective intrusion detection at an early stage 3) Execute autonomous corrective control action by smart devices. Furthermore, we propose to investigate the optimal placement of measurement devices (like PMUs) within the network, to maximize the system’s robustness.