Research On Heterogeneous Software Deployment Strategy For Mimic Defense System

With the continuous integration of network security and new technologies such as big data,mobile Internet,and cloud computing,network attack technology is gradually developing towards intelligence,automation,and diversification.Cyberspace security defenses have gradually changed from traditional defenses that rely on feature detection to the status that need to bring their own “endogenous safety immunity” function.As a new type of active defense technology that changes the rules of cyberspace security games,mimic defense technology has passed many professional system tests and engineering practices,but it still has the problems of limited scope of heterogeneity measurement model and single software deployment strategy.To solve the above problems,this thesis will follow the technical route of software heterogeneity measurement,heterogeneous software deployment,and deployment strategy security gain evaluation to carry out theoretical research,and then combine the security problems of mimic voting software to develop the practical application of theoretical research,so as to provide a set of systematic theoretical practical solutions for the mimic defense system.The main research contents and innovations are as follows:1.A software heterogeneity measurement model for mimic defense system is proposed,which solves the problem that the existing heterogeneity measurement model is difficult to be applied to the software generated by source code compilation.The security of the application software in the mimic defense system depends on two aspects,including the possibility of a single software being attacked and the inconsistency of the structural characteristics of multiple software.This thesis summarizes it as complexity and difference which are known as the characteristic of software heterogeneity,and in this way to propose a heterogeneity measurement model based on software complexity and difference.2.A software deployment strategy based on heterogeneity and performance is proposed to solve the problem of single strategy for heterogeneous software deployment in mimic defense systems.With the heterogeneous and redundant deployment of software in the mimic defense system,the attack chain for the system software changes from available to unreachable or invalid.This thesis summarizes it as heterogeneous software deployment behavior within the mimic defense system.In order to optimize the software deployment benefits in the mimic defense system,the concepts of heterogeneous cost-effectiveness ratio and graph heterogeneity are introduced,the heterogeneous benefits of diverse compilation algorithms are evaluated,and the optimal algorithm is selected to support the research on heterogeneous software deployment strategies.Experiments show that the number of heterogeneous software consumed by the deployment strategy proposed in this thesis is reduced by a maximum of 10.33 times compared to the greedy algorithm.At the same time,according to the research results,a safe operation architecture of the mimic voting software is designed,which improves the security of the voting software in the mimic defense system.3.A security gain evaluation method for heterogeneous software deployment strategies is proposed,which improves the theoretical research on quantitative assessment of deployment strategy benefits.The deployment strategy security gain is an intuitive manifestation of strategic benefits and a determinant of the security of mimic defense systems.The study of its evaluation methods is of great significance.Therefore,the concept of variant common-mode escape probability is introduced to support the study of deployment strategy safety gain evaluation methods.Experiments show that the probability of common mode escape of the system using the deployment strategy in this thesis is reduced by a maximum of 225 times compared to the isomorphic case.