Vulnerability Analysis Method Based On Network Structure And Node Attributes

In recent years,the network security problems are increasing,which leads to the control and management of the network facing more and more serious situation.In order to deal with the threat of network insecurity,network managers are in urgent need of an effective means to strengthen the network security control and management.Network vulnerability analysis is an important part of the prevention of network attacks,which can help network managers to establish complete and effective network control and management mechanism,and is also one of the hot research issues in the field of network science.High-risk nodes in the network refer to the nodes with high influence but low defense.These nodes are the important factors that cause the network vulnerability.Existing identification methods of high risk nodes,such as maximum degree identification and proximity centrality identification,still have some limitations.First,using a single importance parameter,the mined high risk nodes only have influence on the local network,and have little influence on the global network.Second,most of the methods only analyze from the structural aspect and ignore the influence of the node defense capability on the vulnerability analysis.Third,most focus on static network analysis,in the face of continuous network threats lack of dynamic analysis ability.In view of the limitations of existing methods,this theis proposes a network vulnerability analysis method based on network characteristics.The main contributions and innovations are as follows:1.Aiming at the problem of network structure vulnerability,a network vulnerability analysis method based on structural characteristics is proposed.The method uses the minimum input theory to identify the source nodes of information transmission in the network structure by means of network matching.In addition,two methods of node importance quantification,intermediate centrality and K-shell decomposition,were adopted to quantify node importance from the speed and breadth of information transmission respectively.Finally,the high-influence nodes are identified by combining the identification of initial nodes of information transmission and the ranking of node importance.2.Aiming at the problem that it is difficult to integrate network structural characteristics and node attribute characteristics,a vulnerability analysis method combining node attribute characteristics is proposed.This method uses generative antagonism in artificial intelligence to simulate the process of network attack.The generator is responsible for information fusion of the structural vulnerability parameters and the parameters of the node's defense against network attack,and mining the high-risk node set in the network.The discriminant is responsible for simulating the network attack and returning the remaining topology after the network was attacked to the generator for generating the next round of confrontation process.The high risk nodes in the process of information transmission can be dynamically identified by the generator and discriminant.In order to verify the effectiveness of the method proposed in this theis,this theis uses the information transmission model to carry out the information transmission experiment and the network simulation software NS3 to simulate the network attack experiment.Through information dissemination eventually reach the effect of velocity and coverage,and network vulnerability analysis method based on the network characteristics of weak nodes by network intrusion on the influence degree of the network,prove that the method of this theis can effectively identify the high-risk nodes in network,the safe and effective network control and management has important significance.