Research On System Optimal Defense Strategy Based On Dynamic Attack-Defense Game

With the quick growth of information technology,the network environment is getting more and more complicated.Threats and confrontations in network systems lead to increasingly severe security situation in network systems.Research shows that network attack and defense has become the main form of confrontation in all aspects of network systems.The key to guaranteeing the security of network systems is to develop valid protection strategies.In order to effectively deal with the risk of deliberate attacks faced by the system,this thesis studies the optimal defense strategy of the system by constructing a dynamic attack-defense game model.The constructed model comprehensively analyzes the different defense measures taken by the defender,and minimizes the probability of the system being destroyed when the defender has limited resources.The attack-defense game model constructed in this paper is used to describe the game process between the attacker and the defender and to conduct attack behavior prediction and optimal defense strategy research.This thesis deals with two systems in series and in parallel,including four defensive measures: protecting genuine targets,deploying false targets,setting up disinformation and taking preventive strike.This paper mainly studies the optimal defense strategy of the system in the case of introducing intelligent attack and defense and deploying false targets in series system,preventive strike in series system,and preventive strike in parallel system.In this two-stage dynamic attack-defense game with complete information,the defender acts first,and the attacker adopts the attack strategy which can make the system suffer the greatest damage according to the defender's strategy.Since the defender knows that the attacker will take the attack means to maximize the system damage in the second stage,this paper adopts the backward induction method when studying the optimal defense strategy of the defender.When studying the optimal strategy of defender based on dynamic attack-defense game model,the model of attack-defense game is modeled firstly,then the optimal attack strategy of attacker is determined based on the model constructed,then the optimal active defense strategy of defender is determined by minimax algorithm.Finally,the sensitivity analysis of key parameters in the model is carried out by numerical method,which provides a reference for the decision-making of defenders.Compared with the existing attack-defense game model,this paper combines the complexity of the environment in which the current network system is located and the intelligence of the attack mode to construct the following model: a series system attack-defense game model which intellectualizes attack-defense and deploys false targets,a series system attack-defense game model which combines preventive strike by defender with false bases by attacker,and a parallel system attack-defense game model which combines preventive strike with false bases by attacker.The aim is to expand the modeling method of the existing attack-defense game model and improve the accuracy of the system defense theory for analyzing practical problems.This paper studies the series of system that combine intelligent offense and defense with the deployment of false targets,the optimal defense strategy of series and parallel systems that combines preventive strike and false bases.It is hopeful to enrich the research techniques and methods in the field of system defense and provide some reference for the research in related fields.