The feasibility study on detection of unauthorized access into industrial control systems network

We conducted a feasibility study on intrusion detection systems as a cyber-security measure for an industrial control systems network. Intrusion detection systems can detect network connections of malicious terminals and network communication by malicious programs such as computer viruses. We made a comparison with three detection methods and three detection algorithms on control systems availability, detecting functions, ease of deployment and operability. We selected the port mirroring method with machine learning algorithm through the comparison. We introduced an intrusion detection system with the selected method / algorithm on a trial basis to the actual control systems network and verified the functionality. (author)