Anomaly Detection on the HIL-based Augmented ICS Data set

Detecting cyber attacks or abnormal events case using HAI data set

Cyber security risks to national infrastructure and industrial facilities have been steadily increasing in recent months. This is because , over time more control systems are often connected to other systems through the internet, which increases the potential for cyber security threats to the control systems of critical infrastructure. Countries all over the world s tarted to pay attention to the need of protecting Industrial Control Systems ICS. More over , more researchers are directed to begin working to develop security solutions in response to cyber-attacks that have the potential to wreak irreversible damage to countries and society by compromising critical national infrastructure.According to packetlabs during the first half of 2021, 33.8% of ICS computers were attacked, 0.4% higher than in Q2 2020[1].

Number of malware families blocked on ICS computers source [Kerspesky ICS]

To build a machine learning model that effectively deters an attack attempt, it’s crucial for the model to be able effectively to some extent to tell the difference between abnormal system activity and regular states.

The data set we using came from by National Security Technology Research Center which uses GE, Emerson, and Siemens industrial control devices, sensors, and actuators.A control system test bed was built, and the…