Research On Technologies For Network Security Threat Situation Assessment And Analysis

With the rapid development of the network informatization and the continuous improvement of the network technology, various types of confidential information have existed in the Internet. In tradition's network security technique, the method to deal with network attack is to adopt passive defense means which seems not very capable to handle the network attacks which are always changing. In order to make the network security defense system dynamic and change the defense measure actively but not passively, this thesis, on the combination of standards and present situation of network security evaluation, applies the system engineering theory and method to recognizes, analyzes and evaluates the possibility of the inherent or potential threat and risk indicators of the network information security, discusses the steps and process of evaluation of the network security threat, establishes a index system of he network security threat and gives a fuzzy comprehensive evaluation model and relevant evaluation method for network security.This thesis analyzes the main content about network security threat evaluation including obtains a relatively reasonable threat index system and establishes a threat evaluation data model on the basis of the index system. This thesis proposes a weight determination method which combines improved grey correlation analysis and AHP for the system in which indices and targets are quantified. This method avoids that the objectivity of the index weight affects the targets. It also avoids information loss. It eliminates artificial error to a maximum extent by improving AHP.The thesis emphatically brings forward a method to establish a model of the network threat evaluation. This method is based on grey fuzzy weight matrix. By this matrix, the result of the network threat evaluation and a situation chart of the network threat can be obtained. By studying the network threat evaluation data using grey relational analysis, we can get the sorting of every threat type, the sort of threat degree of every host and the corresponding curve graph of the threat status. The method penetrates every phase of the network trend perception, it is the actual the operation object of the evaluation, prediction and the visualization, Whether it is established reasonably or not affects the trend perception result to a great extent. This method through experiment proves the scientific evaluation of the network threat degree is an import link in network security and is the main basis for the active defense of the supervision system. It is also one of the import functions of network security evaluation system. Therefore, it is quite necessary to establish a credible and general index system model of the network threat evaluation.Finally, prove the validity of this method through experiment and present the disadvantage of the method now and the prospect for the risk assessment's development.