Coarse Tree Algorithm-Based Detection of Unstructured Cyber-Attacks in Oil and Gas Process Control Networks

Recently, there have been malicious cyber-attacks on oil and gas installations, which have increased the rate of catastrophic damages leading to loss of containment, loss of life, and destruction of the environment. Cyber-attacks on process control networks could be in different forms, and the inability to detect such unplanned attacks poses a massive problem to this critical infrastructure supplying energy to the world. This paper reviews the different attack methods in SCADA systems of oil and gas installations, proposing the Coarse Tree Machine Learning algorithm to enhance network surveillance and monitoring while protecting the Process Control Network (PCN) data exchange between the sensors, controllers, processes, and the final control element. The experimental results demonstrate the significance, reliability, and applicability of the proposed Coarse Tree algorithm on two different SCADA datasets. The quality and precision of attack detection further validate the proposed model's performance in reducing false alarm rates recorded for the evaluated large SCADA dataset over an extended period.