Network and Risk Modeling for Disaster Survivability Analysis of Backbone Optical Communication Networks

In recent years, survivability of optical communication networks against large-scale disaster failures has been studied. The survivability analyses done by most of the studies are based on arbitrary simplified disaster areas under the assumption of uniform disaster occurrence in a given area, and short-term evaluation of network robustness. However, for assessing the generic applicability and evaluating the performance of any protection, recovery, and/or topology design scheme irrespective of the varying geographical region and network topology, a stochastic model for disaster occurrence and optical network topology would be useful. In this paper, we propose a stochastic model to estimate the impact of earthquake disasters on a backbone optical network, namely, earthquake risk and backbone optical network (ERBON) model. We consider various stochastic distributions, real statistics, stochastic geometry, and graph theory to model seismic zonation, epicenter location and density, earthquake magnitudes, link/node failures, and network topology/connectivity. Furthermore, we propose an earthquake risk minimized node relocation (ERMNR) scheme that can improve the optical network's survivability. Performance of the proposed ERMNR scheme is evaluated under the proposed ERBON model and for real-world earthquake risk data from the U.S. and India.