Assessing the Vulnerability of Network Topologies under Large-Scale Regional Failures

Wei Peng,Zimu Li,Yujing Liu,Jinshu Su

DOI: https://doi.org/10.1109/jcn.2012.6292252

2012-01-01

Journal of Communications and Networks

Abstract:Natural disasters often lead to regional failures that can cause network nodes and links co-located in a large geographical area to fail. Novel approaches are required to assess the network vulnerability under such regional failures. In this paper, we investigate the vulnerability of networks by considering the geometric properties of regional failures and network nodes. To evaluate the criticality of node locations and determine the critical areas in a network, we propose the concept of α-critical-distance with a given failure impact ratio α, and we formulate two optimization problems based on the concept. By analyzing the geometric properties of the problems, we show that although finding critical nodes or links in a pure graph is a NP-complete problem, the problem of finding critical areas has polynomial time complexity. We propose two algorithms to deal with these problems and analyze their time complexities. Using real city-level Internet topology data, we conducted experiments to compute the α-critical-distances for different networks. The computational results demonstrate the differences in vulnerability of different networks. The results also indicate that the critical area of a network can be estimated by limiting failure centers on the locations of network nodes. Additionally, we find that with the same impact ratio α, the topologies examined have larger α-critical-distances when the network performance is measured using the giant component size instead of the other two metrics. Similar results are obtained when the network performance is measured using the average two terminal reliability and the network efficiency, although computation of the former entails less time complexity than that of the latter.