Characterizing and aggregating attack graph-based security metric

An attack graph is an abstraction that represents the ways an attacker can violate a security policy by leveraging interdependencies among discovered vulnerabilities. Attack graph analyses that extract security-relevant information from the attack graph are referred to as attack graph-based security metrics. Although a number of attack graph-based security metrics have been proposed in the literature, there has been no analysis of how these security metrics behave in response to security incidents. In this dissertation, we examine how attack graph-based security metrics behave in response to increased network vulnerabilities under heterogeneous network models. From this analysis, we identify opportunities for using equations that characterize particular attack graph-based security metrics avoiding the costly processing of attack graphs. Security is recognized to be a multidimensional entity. However, all proposed attack graph-based security metrics have been unidimensional. In this dissertation, we provide an approach for aggregating the capabilities of existing attack graph-based security metrics with our proposed suite of attack graph-based security metrics. Lastly, we specify an algorithm for network hardening given a limited budget. Given a set of network vulnerabilities and a set of candidate countermeasures to implement, a network administrator is to choose the set of countermeasures that optimize security given a limited budget. Our algorithm produces sets of counter- measures that optimize security with respect to a set of attack graph-based security metrics while staying within budget