The events of 11 September 2001 brought an increased focus on security in the United States and specifically the protection of critical infrastructure. Critical infrastructure encompasses a wide array of physical assets such as the electric power grid, telecommunications, oil and gas pipelines, transportation networks and computer data networks. This paper will focus on computer data networks and the spatial implications of their susceptibility to targeted attacks. Utilising a database of national data carriers, simulations will be run to determine the repercussions of targeted attacks and what the relative merits of different methods of identifying critical nodes are. This analysis will include comparison of current methods employed in vulnerability analysis with spatially constructed methods incorporating regional and distance variables. In addition to vulnerability analysis a method will be proposed to analyse the fusion of physical and logical networks, and will discuss what new avenues this approach reveals. The analysis concludes that spatial information networks are vulnerable to targeted attacks and algorithms based on distance metrics do a better job of identifying critical nodes than classic accessibility indexes. The results of the analysis are placed in the context of public policy posing the question do private infrastructure owners have sufficient incentives to remedy vulnerabilities in critical networks.

No abstract