MODELING INOPERABILITY PROPAGATION USING BAYESIAN NETWORKS

The modeling of critical infrastructure interdependencies is a challenging task. This paper discusses several interdependency modeling requirements and proposes a Bayesian network approach for modeling interdependencies and inoperability propagation. The approach is applied to a case study involving the Japanese critical infrastructure sectors. Survey data published by the National Institute of Land and Infrastructure Management and the Japanese National Information Security Center are used to generate conditional probability values for the Bayesian network. The approach has the flexibility to adapt to diverse critical infrastructure scenarios and interdependency structures.