Method of Security Situation Assessment Based on Graph Clustering
Wei Yong · Jisuanji gongcheng · 2009
This paper analyzes and compares several typical existing methods of security situational assessment, and proposes a brand new way of assessing the security situation of distributed systems based on network topology, which incorporates both graph algorithms and a clustering algorithm of data mining. Clustering result of the connected graph corresponding to the distributed system is generated by graph clustering algorithm to calculate the integrity reference as a measure of the degree of integrity of the distributed system topology, which is used for quantitative situation assessment. This method achieves less artificial interference and its correctness and applicability are verified by simulated experiments with the network simulation tool.