Security Assessment System Based On Bayesian Network

After the firewall and the intrusion detection, network security assessment is a hot issue ready to be focused on in network defensive field. It helps us to analyze and assess the security status of computer network system; it also assists us to figure out and deal with vulnerabilities which exist in the system, thereby greatly reducing the security risk of a system and guaranteeing a network system to work normally. In the process, network security assessment is used to make certain the risk level and provides effective approach to protect network information.This paper first analyzes the correlation among different vulnerabilities and integrates large amount of messages from multiple vulnerability database. Then a quantitative vulnerability database comprising 30 thousands of vulnerabilities is being built, based on which the writer proposes an attack graph generation approach according to atomic domain, while cutting short the process of generating attack graphs. Then the evaluative model of Bayesian attack graphs is constructed and a fresh evaluative method is put forward grounded on the previous work. By automatically assign probability information to a Bayesian attack graph, this method actually achieves the automation of evaluation, thus applicable to large-scale networks. All the above work contributes to design a new evaluation system and engaging testing and analysis of the system capability in a network environment. The results of the study demonstrate that this evaluation system provides a scientific evaluating basis for a network administrator. This has finally been proved to be a necessary method to improve network security.