Research On Attack-defense Game In Complex Networks Under Asymmetric Conditions

Under the background of the attack of UAV cluster on the infrastructure system,how to take advantage of asymmetric information to make active defense and allocate limited defense resources efficiently and expediently is an important research topic for intelligent decision-making in intelligent combat in the future.On the one hand,as combat elements become increasingly diverse and their interrelationships become increasingly complex,it is necessary to conduct concise and efficient modeling so as to quickly grasp the core characteristics of such complex systems from complex battlefield data and provide data support for intelligent decision-making.On the other hand,the defense decision maker needs to convert the information advantage into decision advantage in the case of asymmetric information advantage,and carry out active defense against the attack of UAV cluster.Based on these two goals,the main work and innovations of this paper are as follows:(1)In view of the reality of asymmetric information condition,a method of constructing false network is used to construct asymmetric information condition for the first time on the basis of complex network modeling.In the modern war,the infrastructure system is complex,and the information of attack and defense is asymmetric.In this paper,an active defense method using the advantage of asymmetric information is provided to expand the attack and defense game model under the condition of asymmetric information.The false information is constructed by the method of random adding,deleting and connecting edges,which keeps part of the topological properties of the network and makes the false network both effective and deceptive.(2)Aiming at the fact that the defensive side deploys the defense strategy before the attacking side,and the attacking side chooses its own attack strategy according to intelligence,in this study the asymmetric information condition for the game target is combined with Stackelberg safe game for the first time.The dynamic game of Active Deception Defense(ADD)based on Stackelberg security game is studied.The defender is able to take advantage of the asymmetric information while gaining the first-mover advantage.When the noise level α = 0.1 and 0.5,the average payoff of the defender increased by 4.7% and 13.59%,respectively.(3)In view of the complex situation of multiple types of attackers,the active spoofing defense dynamic game under asymmetric information condition is extended to bayesian dynamic game for the first time.In this study,the prior probability of the occurrence of multiple types of attackers is taken as input to provide a computing framework for quick decision-making in complex environments facing multiple types of attackers,and the sensitivity of the prior probability is analyzed.When the defender is faced with multiple types of attackers,overestimating the prior probability of an attacker with a high risk is more beneficial to defenders than underestimating the prior probability.(4)Aiming at the problem of active deceptive defense of infrastructure system under the condition of asymmetric information,this paper constructs the scenario of the infrastructure system being attacked by UAV cluster to test and verify the active deceptive defense method.Based on the background that the infrastructure network is attacked by UAV cluster,the experiment respectively constructs the scenario that the infrastructure network conducts active deception defense against UAV cluster attack and the scenario that the UAV cluster conducts active spoofing defense against ground air defense weapon attack.In this study,the real power system is modeled by a single-layer network,and the UAV cluster system is modeled by a multi-layer network,providing a model support for the battlefield decision maker to grasp the system core quickly.The experiment proves the effectiveness of the active deception defense method in transforming information advantage into decision advantage.