Vulnerability Assessment of Community-Interdependent Infrastructure Network Based on PSDA

Abstract

Vulnerability assessment of interdependent infrastructure networks (IINs) must be conducted according to a specific situation such as the failure model, the scale of networks, interdependent relationships, infrastructure characteristics, and so on. This paper discusses the development of an interdependent network model for vulnerability assessment by combining the characteristics of community IINs and the design of the maximum flow comprehensive index (MFCI) to describe the performance of interdependent networks based on network flow theory. This study is based on the perspective of an infrastructure manager, and identifies the key nodes of interdependent networks by using the optimal attack strategy when dealing with potential external attacks (i.e., terrorism). It is assumed that attackers are sophisticated and will always choose the optimal attack strategy to maximize vulnerability and minimize the performance of interdependent networks. Because it is difficult to identify key nodes in interdependent networks, we proposed a probabilistic solution discovery algorithm (PSDA) that can accurately calculate the final failure state of interdependent networks regarding the interdependent relationships among networks and relationships inside a network. A case study was conducted to validate the proposed method: using MFCI and PSDA. Moreover, the assessment result was analyzed further with a view to improving infrastructure protection and identifying failure propagation paths. The main contributions of this paper are to propose a vulnerability assessment index system and construct a goal optimization with solving algorithm to provide decision support for the infrastructure managers.