A Bayesian Network Method Of Tracing Cyber Attacks

Along with the popularity of the Internet comes various forms of network attacks and cyber crime, and they have caused an extremely seriously affect to people's life and work. Because of the popularization of the network, more people have access to relevant knowledge of networks, which makes the means and methods of cyber attack and defense become more diverse and more specialized. Scholars and network-related staff all over the world are working to prevent and track malicious and non malicious program errors caused by network attacks and have made great progress. However, these methods are based on the network itself, this paper presents a Bayesian network using mathematical model combined with the traditional model to locate a cyber attack. The main contribution of this paper:1) Analysis of the traditional network attack tracing methods and apply Bayesian network structure which is a mathematical model of statistical methods to solve the problem of locating interruption caused by lacking of evidence;2) Through the data fusion method the paper analyzes past data of cyber attacks and find out the potential regular without consuming network resource:3) The application of Bayesian network model establishes a global observe system which could be used by daily maintenance of network equipment and evaluation.