SSO-IF: An Outlier Detection Approach for Intrusion Detection in SCADA Systems

Supervisory Control and Data Acquisition (SCADA) systems play a prominent role in monitoring and controlling the Critical Infrastructures (CIs) such as water distribution, nuclear plants, and chemical industries. On the other hand, SCADA systems are highly exposed to new vulnerabilities as it highly relies on the internet. Machine learning approaches have been employed to detect the cyberattacks injected by the attackers in CIs. However, those approaches failed to protect the CIs against the ever-advancing nature of cyberattacks. This work presents Salp Swarm Optimization-based Isolation Forest (SSO-IF) to build an efficient SCADA intrusion detection system, and the experiments were carried out using power system dataset from Mississippi State University. The performance of SSO-IF was validated over the state-of-the-art intrusion detection techniques in terms of classification accuracy and detection rate.