Critical infrastructure networks are becoming increasingly interdependent. An attack or disaster in a network or on a single node in a network will affect the other networks dependent on it. Therefore, it is important to assess and understand the vulnerability of interdependent networks in the presence of natural disasters and malicious attacks that lead to cascading failures. We develop a framework to analyse the robustness of interdependent networks. Nodes and links in the interdependent networks are attacked based on the graph centrality metrics. We apply our framework on critical infrastructure network data.Our results indicate that the importance of critical infrastructure varies depending on the attack strategy.

No abstract