A General Framework for Network Survivability Quantification

In this paper, we propose a general survivability quantification framework which is applicable to a wide range of system architectures, applications, failure/recovery behaviors, and desired metrics. We show how this framework can be used to derive survivability measures based on different definitions and extend it to other measures not covered by current definitions which can provide helpful information for better understanding of system steady state and transient behaviors under failures/attacks. An illustrative example of a telecommunications switching system is given for the ease of discussion. Markov models are developed and solved to depict various aspects of system survivability.