Last modified: Friday, April 13, 2007
Planning for threats to critical infrastructure networks
National security requires new approach, IU expert says
FOR IMMEDIATE RELEASE
April 16, 2007
BLOOMINGTON, Ind. -- Violent storms and terrorist attacks represent two different kinds of threats to the nation that Department of Homeland Security staff must anticipate. And in both of those scenarios, planners must stay one step ahead of threats to critical infrastructure networks.
Tony Grubesic, assistant professor of geography at Indiana University Bloomington, studies the consequences of infrastructure failure and disruption of networked systems such as the interstate highway system, telecommunication backbones, commercial airline networks and the electrical grid.
"Basically, I'm developing methods to evaluate, monitor and mitigate the impacts of extreme events to critical infrastructure systems," Grubesic said.
In a chapter that appears in a forthcoming book that he co-edited, Critical Infrastructure: Reliability and Vulnerability (Springer-Verlag), he and his co-authors evaluate the consequences of losing "vital nodes" in geographically-linked networks, such as telecommunication switching centers or electrical substations.
Their method uses mathematical programming and geographic-information systems to determine the most efficient scenarios for disabling infrastructure systems.
This type of analysis is important for several reasons, Grubesic said. It draws attention to the strengths and weaknesses of the various network structures used for infrastructure design, highlighting the geographic propagation of failure through systems. It also provides a foundation from which better strategic decisions in network design can be made.
"In addition, we also can begin to estimate the population at risk for exposure to extreme events and catastrophic failures," he said. "For example, by calculating the number of consumers or corporations that are dependent on critical infrastructure systems, and where they are located, we can provide a window for estimating the potential economic, social or cultural impacts that a disruption may cause."
Much of the work being done in this field focuses on system connectivity as the most important factor for examining network disruption, Grubesic said. He feels that this approach ignores other factors that are at least as important.
"Our research suggests that the way in which networks are used is more important," he said. "For example, consider a single node on a network, such as a telecommunication hotel. This is a commercial floor space that provides interconnection for competing networks, as well as operational, administrative and management interfaces.
"This telecommunication hotel might have hundreds of connective links to other telecommunication hotels, networks or systems," Grubesic continued. "However, the use of the fiber optic links connecting these nodes can vary greatly -- as can the geographic distribution and arrangement of these systems. As a result, attacks on highly connected elements of the system will not always yield the largest disruption."
Instead, he said, it might make more sense to target a somewhat less connected element that experiences heavier traffic loads.
"In reality, the optimal attack would be a combination of highly connected and/or highly used locations on the network. Finding these combinations out of millions of possibilities is the trick," he said.
Once these potential targets have been identified, then planners can examine strategies for improving the survivability of critical infrastructure systems, Grubesic said. In some cases, for example, this might include fortification of important facilities that are particularly vulnerable. In others, it might simply mean adding a few links to the system or increasing capacity along existing links to provide alternatives for the delivery of goods and services when systems are disrupted.
"Regardless of the strategy ultimately selected, the economic, political and social ramifications of failing to address these problems could cost the U.S. economy billions," he said.