Date of Award
Doctor of Philosophy (PhD)
Industrial Control Systems (ICSs) are designed, implemented, and deployed in most major spheres of production, business, and entertainment. ICSs are commonly split into two subsystems - Programmable Logic Controllers (PLCs) and Supervisory Control And Data Acquisition (SCADA) systems - to achieve high safety, allow engineers to observe states of an ICS, and perform various configuration updates. Before wide adoption of the Internet, ICSs used "air-gap" security measures, where the ICS network was isolated from other networks, including the Internet, by a physical disconnect . This level of security allowed ICS protocol designers to concentrate on the availability and safety of operation of physical systems while decreasing the need for many cyber security implementations. As the price of networking devices fell, and the Internet received global adoption, many businesses became interested in the benefits of attaching ICSs to wide and global area networks. However, since ICS network protocols were originally designed for an air-gapped environment, it did not include any of the security measures needed for a proper operation of a critical protocol that exposes its packets to the Internet.
This dissertation designs, implements, and evaluates a telemetry based Intrusion Detection System (IDS). The designed IDS utilizes aggregation and analysis of the traffic telemetry features to classify the incoming packets as malicious or benign. An IDS that uses network telemetry was created, and it achieved a high classification accuracy, protecting nodes from malicious traffic. Such an IDS is not vulnerable to address or encryption spoofings, as it does not utilize the content of the packets to differentiate between malicious and benign traffic. The IDS uses features of timing and network sessions to determine whether the machine that sent a particular packet and its software is, in fact, a combination that is benign, as well as whether or not it resides on a network that is benign. The results of the experiments conducted for this dissertation establish that such system is possible to create and use in an environment of ICS networks. Several features are recognized and selected as means for fingerprinting the hardware and software characteristics of the SCADA system that can be used in pair with machine learning algorithms to allow for a high accuracy detection of intrusions into the ICS network. The results showed a classification accuracy of at least 95% is possible, and as the differences between machines increase, the accuracy increases too.
Ponomarev, Stanislav, "" (2015). Dissertation. 199.