Measuring Network Vulnerability Based on Pathology

This paper compares disease with network vulnerability by their definitions and characteristics. A mapping between disease and vulnerability is built based on their similarities. We put forward a novel model of vulnerabilities in computer networks by simulating the reverse of cause-result of disease. Based on the model, a quantitative metric for vulnerabilities of computer networks is presented. The complexity of the algorithm for computing the metric is O(|V|²X|S|), where V and S stand for set of vulnerabilities and set of network states. By analyzing different structures of the vulnerability model, we found that the value reflecting vulnerability decreases when the model is more linear.