In addition to the research in the area of wireless networking, we have also been studying the resilience of telecommunication and power networks to large scale geographically correlated failures. Since these networks rely on physical infrastructure, they are vulnerable to natural disasters, such as earthquakes, hurricanes, floods, and solar flares, or to physical attacks, such as an electromagnetic pulse (EMP) attack. Hence, we focus on the vulnerability of these networks to failures in a common geographical area (i.e., to geographically-correlated failures). Developing tools for identifying vulnerabilities is of utmost importance for network monitoring, strengthening, and modernization.
We studied the effects of deterministic and probabilistic geographically correlated failures on communication networks. Under the deterministic model, all lines in a specific region fail and under the probabilistic model, the failure probability of a line is a function of its distance from the event epicenter (since the effects of physical attacks can rarely be determined in advance, probabilistic models are more realistic). We also considered scenarios with a number of simultaneous attacks and in which a protection plan is in place. We developed efficient algorithms that find a worst-case attack location and obtained numerical results for backbone networks, thereby demonstrating the applicability to real-world networks.
Moreover, we considered the power grid vulnerability and studied the unique effects of geographically correlated outages. We developed tools for identifying the most vulnerable locations in the grid and performed extensive numerical experiments with grid data to investigate the various effects of geographically correlated outages and the resulting cascades. These results allowed us to gain insights into the relationships between various parameters and performance metrics, such as the size of the original event, the final number of connected components, and the fraction of demand (load) satisfied after the cascade. We also considered various computational aspects of cascading failures in power grids and showed the limitations of epidemic- and percolation-based tools in modeling the cascade evolution.
In our recent work we have been considering the vulnerability of the power grid to joint cyber and physical attacks and developing methods to retrieve the grid state information following such an attack. Finally, we have been studying the structural properties of the North American grids and developing algorithms for generating synthetic power grids (i.e., spatially embedded networks with similar properties to a given grid). This work is motivated by the fact that the development of algorithms for enhancing the resilience of the power grid requires evaluation with topologies of real transmission networks but such topologies are usually not publicly available.
The video below includes a short talk about the “vulnerability of power grids to geographically correlated failures” that was given in the FCC Workshop on Network Resiliency in Feb. 2013.
P. K. Agarwal, A. Efrat, S. K. Ganjugunte, D. Hay, S. Sankararaman, and G. Zussman, “The resilience of WDM networks to probabilistic geographical failures,” IEEE/ACM Transactions on Networking, vol. 21, no. 5, pp. 1525–1538, Oct. 2013.
A. Bernstein, D. Bienstock, D. Hay, M. Uzunoglu, and G. Zussman, “Sensitivity analysis of the power grid vulnerability to large-scale cascading failures,” ACM SIGMETRICS Performance Evaluation Review, vol. 40, no. 3, pp. 33–37, Dec. 2012.
S. Neumayer, G. Zussman, R. Cohen, and E. Modiano, “Assessing the vulnerability of the fiber infrastructure to disasters,” IEEE/ACM Transactions on Networking, vol. 19, no. 6, pp. 1610–1623, Dec. 2011.