Research On Network Attack Path Prediction Method Based On T＿NAG Model

With the wide application and rapid development of Internet technology,the number of computer malicious programs continues to grow at a high speed,and the security incidents caused by network attacks are endless.How to predict the occurrence of network attacks and update the network security strategy according to the prediction results in time to avoid unnecessary losses as much as possible is the main task of network security researchers at this stage.Based on the study of the existing network attack prediction methods,a new attack graph model T＿NAG is proposed.In addition to the conventional factors,the time attribute of the attack behavior and the dynamic characteristics of the attacker’s ability are fully considered in the analysis of the attack intention.Using the advantage of attack graph model,the reachability probability of each path is calculated from the local node reachability probability analysis,and the security of the whole network is predicted.The main contents of this paper are as follows:1)A time attribute network attack graph model is designed.Based on the study of the theory of attack graph and the general law of attack behavior,the generation algorithm At＿Aggalg of the model is proposed.2)According to the attacker’s real-time behavior track to distinguish the attacker’s ability,this paper proposes an attack intention analysis model based on the attacker’s ability,considers the operation risk and attack benefit as a whole,introduces the time decay parameter into the attack intention calculation,and designs a vulnerability utilization quantification method based on the attacker’s ability.3)In order to improve the accuracy of network attack behavior prediction,the intention of attackers and vulnerability utilization rate are further integrated into the consideration of path reachability probability.Based on Bayesian theory,a quantitative method for each node and path is designed.Finally,the attack path is predicted by using the proposed Inten Abi-Path Pre algorithm.