Research On Network Security Risk Assessment Method Based On Bayesian Attack Graph

With the development of the Internet,cyber security risks are increasing.New types of emerging network attacks pose new challenges to the smooth and efficient operation of the network.Among them,the network security risk assessment mistake,which caused by the inaccurate correlation analysis between attack intention and system defects,and the failure of network security protection measures,which caused by unbalanced measures costs and benefits,have become the key issues in the field of network security.An effective method is to balance the costs and benefits of the protection strategy by establishing a risk assessment model,which analysis the correlation between system vulnerability and attack intention.Based on this,we carry out a systematic study of risk assessment method from the perspective of risk analysis and risk management.The main work of this paper are as follows: 1.By studying traditional network security risk assessment techniques,we point out the design defect of risk analysis methods and protection strategy selection methods.Combining with attack graph model and intelligent optimization algorithm,a network security risk assessment implementation process that can embed network attack association analysis and protection strategy benefit analysis is designed.It laid a theoretical foundation for the research on cyber security risk assessment methods.2.In order to solve the problem of inaccurate risk assessment results,a cyber security risk analysis method based on Bayesian attack graph is proposed to effectively matching the attacker's intent with system's vulnerability information.The method first calculates the possibility of exploiting vulnerabilities.Secondly,it uses the Bayesian inference method to analyze the impact of network attacks.Finally,an attack graph model for dynamic risk analysis is constructed.Experiments show that the accuracy of network security risk analysis method based on Bayesian attack graph is increased by 26.3% compared with the traditional method,and the accurate analysis of network risk is realized.3.In order to improve the risk management profits,the network protection strategy selection method based on high-return ant colony optimization is proposed.This method first expands the definition of Bayesian attack graph,then combines the extended attack graph and the expected income to predict the attack path.Secondly,using economic indicators quantify the costs and benefits of the network protection strategy.Finally propose a high-return ant colony optimization to balances the costs and benefits of the protection strategy,then obtains the optimal protection strategy set.Experiments show that the ant colony algorithm-based network security protection strategy selection method can reduce the cost by 42.73% while increasing the benefit by 17.29%,and realizes the precise deployment of protection strategy.