College

Bellini College of Artificial Intelligence, Cybersecurity and Computing

Mentor Information

Robert Karam

Description

Industrial Control Systems (ICS) have increasingly been subject to malicious cyber attacks, thus threatening the operational safety, stability, and reliability of critical infrastructure. While existing research has utilized electromagnetic, timing, and audiovisual side channels for ICS attacks, the offensive utility of magnetic side channels in these environments has yet to be explored. Utilizing the magnetometers present in modern-day smartphones, the magnetic fields produced by the Programmable Logic Controller (PLC), Human-Machine Interface (HMI), Linear Actuator Controller, and AC-to-DC Power Converter are recorded. Magnetic field data then undergoes signal processing coupled with machine learning classification on extracted feature vectors to determine the operational states of system components, such as motor activation, linear actuator activation, and linear actuator motion. Preliminary results indicate smartphone-based magnetometers are effective in component-adjacent analysis of ICS operational activity. Findings underscore the need for multi-layer physical security controls within ICS facilities, such as magnetic shielding around PLC technology, to address potential threats posed by mobile magnetometers.

Share

COinS
 

Smartphone-Based Profiling of Industrial Control Systems Using Magnetic Side Channels

Industrial Control Systems (ICS) have increasingly been subject to malicious cyber attacks, thus threatening the operational safety, stability, and reliability of critical infrastructure. While existing research has utilized electromagnetic, timing, and audiovisual side channels for ICS attacks, the offensive utility of magnetic side channels in these environments has yet to be explored. Utilizing the magnetometers present in modern-day smartphones, the magnetic fields produced by the Programmable Logic Controller (PLC), Human-Machine Interface (HMI), Linear Actuator Controller, and AC-to-DC Power Converter are recorded. Magnetic field data then undergoes signal processing coupled with machine learning classification on extracted feature vectors to determine the operational states of system components, such as motor activation, linear actuator activation, and linear actuator motion. Preliminary results indicate smartphone-based magnetometers are effective in component-adjacent analysis of ICS operational activity. Findings underscore the need for multi-layer physical security controls within ICS facilities, such as magnetic shielding around PLC technology, to address potential threats posed by mobile magnetometers.