Detection And Recommend Model Based On Behavior Trust Management

With the development of internet and software technology, applications are movingaway from a closed network environment, being transferred to the open world softwaresystems composed of a multiple cross-domain in internet. In order to detect the largenumber of unknown behaviors and prevent the coordinated attack and improve entitybehavior detection accuracy and reduce the system false alarm rate and false negativerate, the behavior trust management mechanism is introduced into Intrusion PreventionSystem. The IPS provide a reference for the behavior detecting and recommendingthrough building trust relations for the behavior and analysing the history of behaviorinteraction, this paper proposes a research projects for the detection and recommendmodel based on behavior trust management, the main work in the following areas:Firstly, this paper summarized the relevant concepts of Intrusion Prevention System,analyzed the intrusion detection and its shortcomings, described the concept of behaviortrust, then two classic trust model are given. At last the Map/Reduce mechanism, used forthe behavior recommend model, is introduced.Secondly, in order to detect the large number of unknown behaviors and prevent thecoordinated attack, behavior trust management mechanism is introduced to the IntrusionPrevention System to build trust files for the entity behavior, and then the detectionmodel based on behavior trust management is given.Thirdly, the data update algorithm of behavior pattern library is complex and takestoo much time. In order to solve the issues, the behavior trust mechanism is introduced,and a collaborative recommendation algorithm is given to update the trust behavior andharmful behavior to the behavior patten library, improve the speed of detect and behaviordetection accuracy.Finally, simulation shows that the model can effectively reduce the false negativerate and false alarm rate of Intrusion Prevention System, update the characteristics oftrust behavior and harmful behavior to the behavior pattern library quickly, greatlyimprove the behavior detection accuracy, reduced the false negative rate and false alarm rate of Intrusion Prevention System.