Research On Information Security Evaluation Methods For Network Devices

Cyber security issues are more and more concerned by people.Network devices are served as the network infrastructures and the security of them indirectly reflects the network security posture.With the occurrence of large-scale attacks result from the vulnerability of network devices,the problems of network device security gradually come into people's eye.With respect to the lack of standards and indicators of network device information security evaluation,this thesis discusses the security threats of network devices and analyzes the reasons for the inaccurate evaluation of information security evaluation.From the perspective of information security evaluation for network devices,the main contributions of this thesis are summarized as follows:(1)Proposed the information security evaluation framework and evaluation indicators for network devices.On the basis of extensive research,this thesis studies the current situation of information security evaluations and network equipment security technologies,conducts in-depth study on the evaluation criterion and evaluation methods,and analyzes the security threats confronted by network devices.Then,this thesis proposes the relevant information security evaluation framework and evaluation indicators.(2)Proposed a new evaluation method based on D-S evidence theory and principal component analysis.Aiming at the evaluation uncertainty problem,from the perspective of ambiguity,inconsistency and non-specificity,this thesis improves the D-S evidence theory and combines it with principal component analysis.The results show that this method can effectively reduce the uncertainty and subjectivity in the evaluation.(3)Proposed an evaluation method of information security based on device running state.The router,which is the typical representative network device,is taken as an example here.Firstly,the DDoS attack network is built with the help of GNS3,and the data from the CPU utilization,packet loss rate and port flow of routers is monitored and collected in real time.Secondly,the information entropy of the port flow and operation state are analyzed,based on which the classification characteristics are constructed.Finally,on the basis of logistic regression model,the prediction of DDoS is realized.Simulation results show that the accuracy of the proposed method to determine the router in the network attack environment is 91.84%.It can effectively detect the abnormal state of routers and provide a powerful decision basis to the network administrators.From the perspective of information security evaluation,this thesis analyzes the security of network device.The simulation results show that the evaluation indicators and evaluation methods have achieved good results no matter in static or dynamic settings,and have some guidance for the information security evaluation of the network devices.