A Machine Learning Test Data Set for Continuous Security Monitoring of Industrial Control Systems

Continuous security monitoring provides an effective proactive defense against devastating attacks on critical infrastructures such as those operated by industrial control systems (ICS) and Supervisory Control and Data Acquisition (SCADA) system. An essential component of a continuous security monitoring system is an intelligent backend that is built and trained to handle real-time data analytics. A plethora of data sets for intrusion detection systems and network usage facilitated numerous research works on intrusion detection systems and network usage analytics. Although the need for test data sets for ICS security is quite pronounced, there is a discernible deficiency of the availability of such data sets. The contribution of this research is to close that identified gap by generating a test data set using an ICS testbed and to utilize machine learning algorithms for its evaluation.