Research On Network Dynamic Defense Strategies Based On Game Model

With the development of information technology,cyberspace,and human society are deeply integrated,but at the same time,network attacks occur frequently,and the overall network security is at a disadvantage of"easy to attack but difficult to defend".In this context,dynamic defense makes cyberspace the ability to dynamic change and improves the security defense capability of information systems by building and implementing dynamic transfer attack surface diversity strategies.In the deployment process of dynamic defense strategy,the selection of its transfer elements,transfer time,and transfer mode is the basis for improving the security defense capability.Based on this,this thesis has carried out research on network dynamic defense strategies based on the game model.The main work and innovations are as follows:(1)A method of selecting multivariate combination strategies based on the evolutionary gameAiming at the problem of selecting dynamic defense elements when deploying dynamic defense technology in the face of multiple attack environments in information systems,this thesis studies and analyzes the combination deployment selection of dynamic defense elements,and puts forward a method for selecting multiple combination strategies based on the evolutionary game.This method establishes a mathematical model of a dynamic evolutionary game,considers defense cost and defense effect,gives a defense strategy selection algorithm based on an evolutionary game,and realizes the optimal combination of defense elements.Experiments in different system resource environments show that the proposed method is feasible and effective.Compared with the strategy selection algorithm based on the genetic algorithm,it can improve the stability by about 15%and reduce the single run time by about 50ms.(2)Cycle switching strategy selection method based on improved FlipIt gameAiming at the problem of insufficient security defense effect caused by the existing cycle-switching strategy not considering attack intensity,migration time and other factors,a cycle-switching strategy selection method based on the improved FlipIt game is proposed.By extending the FlipIt game model,the defense switching time is taken as defense cost,the measurement unit of defense cost and defense revenue are unified,and attack intensity is set as an attack strategy,the best time for the defender to take the transition is given through the game equilibrium,and the model is consistent with the actual attack and defense environment.Experimental results show that the proposed method can improve the accuracy of dynamic defense timing.(3)A method for selecting time space switching strategy based on the Stackelberg gameAiming at the problem that the space-time strategy is not fully integrated into the deployment of defense strategy when the information system applies dynamic defense technology,this thesis proposes a method to select the space-time combination switching strategy based on the Stackelberg game.This method is based on the dynamic defense attack and defense scenario,combined with the characteristics of network attack and defense,considering that the migration cost and attack time vary with the system configuration,and adopts multiple time intervals based on improved FlipIt game and space selection based on Stackelberg game to realize the combination of space and time decisions in dynamic defense.Through the dynamic defense transformation experiment on the distributed file system,the results show that this method improves the defense benefits,increases the randomness of the transfer,and improves the transfer dynamics through multiple time intervals.