Application Research Of Fuzzy Clustering In Network Running Situation Awareness

Network running situation awareness is about the overall status of the network and possible trends composed by the running situation of network equipments,the use of network resources,and the status of service.Firstly,complete the collection and preprocessing of the data,and then evaluate and predict the network situation according to the evaluation indicators of the network running situation awareness.Finally,it is displayed in a visual interface.The results are presented to network administrators,and help them understand the entire running situation of the network and process the network as accurate as possible.This thesis studies the situation awareness model,establishes the system of evaluation indicators of network running situation awareness,and evaluates the indicators to obtain the running quality of the links,and then integrates the link weights to assess the current running state of the entire network.The specific work is as follows:1.Research the network situation awareness system model.The model of the network running situation awareness system based on the evaluation of link running status is proposed,and various factors are taken as the indicators of situation assessment to realize the assessment and display of the network running situation.2.Study the evaluation indicators of the network running situation.Based on the business,performance,and topology,and taking full account of other factors that can affect the network situation,this thesis establishes the system of evaluation indicators that is required for the study of network situation awareness.3.Study the method of situation assessment.In view of the particularity of network data clustering and the deficiencies of existing clustering algorithms,an improved semisupervised FCM clustering algorithm is proposed that uses Canopy clustering to obtain the initial clustering center,adding marked information and reference membership degree.This algorithm is used to obtain the running situation of each link in the network.Based on the reachability of nodes,a method to calculate the link weights is proposed.And then integrate the status and weight of each link to complete the running situation assessment of the entire network.Based on the existing research on situational awareness,this thesis presents a model of running situation awareness.In this model,a method to assess the operation situation based on the improved semi-supervised FCM clustering algorithm is applied to make assessment.The results of experiments prove that the method can reflect local and overall running situation of network,the results can be used as a reference for the network administrators to manage network.It also brings convenience to the management of the network.