Research On Network Attack Effect Evaluation Model Based On Rough And Fuzzy Comprehensive Evaluation

As a method of proactively discovering the weak links of protection in network systems,network attack effect evaluation has been widely studied and applied in recent years.Among the commonly used models at this stage,the network attack effect evaluation model based on the fuzzy analytic hierarchy process has problems such as strong subjectivity in the index weight calculation process,which leads to unstable and unobjective calculation results of the network attack effect evaluation model using this model.Therefore,it is expected that by optimizing the network attack effect evaluation model based on the fuzzy analytic hierarchy process,the calculation results have higher objectivity and stability,and provide better reference value for network system protection and improvement.Based on a large number of preliminary studies,this thesis studies and designs a network attack effect evaluation model based on rough set and fuzzy comprehensive evaluation for the problem of strong subjectivity of the attack effect evaluation model based on fuzzy analytic hierarchy process,and introduces rough set theory for weighting The calculation solves the problem of strong subjectivity of index weight,and for the problem of the weak ability of the classical rough set theory to deal with incomplete data sets,the K-nearest interpolation method is introduced for data completion and processing.On the premise of not affecting the evaluation results,maintain the changing laws and research value of the data itself;by introducing the above methods,we can obtain more objective and accurate evaluation results of cyber attacks.The main work and innovations are as follows:First of all,basing on the research of a large number of network attack effect evaluation attack index systems,this thesis studies and designs a network attack effect evaluation index system classified by attack purpose.This indicator system removes some indicators which are difficult to collect,reduces indexes of the redundancy,but also ensures that the entire indicator system scalability.Secondly,in the part of determining the weight of the network attack effect evaluation index,the classic rough set theory is introduced to determine the weight.Based on the advantage that it can calculate the index weight without relying on prior knowledge,it effectively improves the objectivity of the evaluation weight of the network attack effect.And the calculation does not require consistency verification,which is simpler than the calculation process of the analytic hierarchy process.Next,in view of the poor processing ability of the classic rough set theory for incomplete data sets,the K-nearest interpolation method is introduced for data completion to improve the robustness of the model.Comparative experiments show that under the premise of a small amount of data,the K-proximity interpolation method has a better completion effect.And this thesis verifies the rationality of the results of the evaluation model of network attack rough sets and fuzzy Comprehensive Evaluation.Next,a feedback mechanism is proposed to circumvent possible problems in the data collection process when the actual model is applied,feedback indicators for not obtaining data,and adjust the evaluation results of network attacks.This method can improve the flexibility and accuracy of the network attack effect evaluation system designed in this thesis.Finally,this thesis designs and implements a network attack effect evaluation system based on rough set and fuzzy comprehensive evaluation.The system is divided into six subsystems,and the overall architecture of the system,the functions of each subsystem and the design and implementation of core modules are described.Passed the designed test cases,and verified the effectiveness and usability of the system functions.