Adapting traditional logic modeling techniques to address cyberattack

This paper is concerned with modeling the impact on systems of attacks that corrupt the flow of information in a system, especially control signals, with a view to taking steps to prevent the adverse impacts of such attacks. “Corrupting the flow of information” includes cyberattack, but is not limited to cyberattack. This scope also goes beyond hardware: it includes corrupting the information made available to operators, in order to induce them to contribute to undesirable plant outcomes. In order to clarify the method, we begin with a simple control-system problem first posed by Lapp and Powers. Even in this simple system, there are several ways to mislead the operators; one way is to induce them to take adverse control actions, and another is to conceal from them an adverse system evolution being induced by corruption of information flow. The paper describes (1) certain adaptations needed to appropriately refocus the application of logic modeling to this class of problems, (2) analysis and tests of attack scenarios, and (3) limitations that need to be addressed in future methodology development. We illustrate Top Event Prevention Analysis in the context of cyberattack; and finally, we discuss future work using a plant simulator.