Research On Improved Trusted Network Connection Based On Behavior Analysis

With the development of network technology and the gradual popularization of lager scale distributed system,network security is facing great challenges.The future of network is the trusted network with the analysis of user behavior,the credibility analysis of user behavior in trusted network has become an important research point in network security.Due to the variability of the network environment,the diversity of the access terminal and the complexity of the user behavior,it is difficult to meet the new requirement of network security in the existing access control mechanisms.Therefore,we need to increase the user behavior trust on the basis of the existing security mechanisms and establish a trust relationship between the users in one platform,the user and the platform,cross platform users,so as to fundamentally solve the problem of network security.Identity authentication and integrity authentication of terminal platforms can be realized by traditional trusted network access control method,which belongs to the static authentication mechanism.However,the authentication results become invalid extremely easily because terminal platform environment and user behavior changing in the actual network environment,which ultimately causers wrong authorizations.In view of the above problems,the essential trusted characteristics of the terminal environments and the user behavior is comprehensive analyzed,then two types of user behavior trust evaluation mechanism and a user behavior network access control model are proposed.The main work in this paper includes the following aspects:(1)A real-time evaluation mechanism based on double evidence classification of user behavior is proposed,which realized the user behavior evaluation,denoted as DEC-UB.The process classification can divide the user behavior user behavior into several time span based on double sliding window,and the characteristic classification can calculate the direct credibility in each time span based on analytic hierarchy process,and the time factor is introduced to update the user behavior credibility,and finally determine whether the user behavior can be trusted by the comprehensive judgment of the credibility.The evaluation mechanism makes the user behavior evidence of any time can be directly involved in the trust evaluation,and the evaluation result is more comprehensive and accurate.(2)We propose a real time evaluation model based on globe trust of user behavior,denoted as RTEM-GT.RTEM-GT use the fuzzy comprehensive evaluation method to get the user behavioral evidence weights and fuzzy evaluation value by considering the fuzziness of network user behavior.The penalty mechanism is introduced to increase the user behavior evaluation method from a single evaluation to a global evaluation.(3)A network access control model based on user behavior trust policy is designed within the trusted network connection framework,and the we put forward the dynamic authorization mechanism of network connection and access.