Research On Virus Transmission Model And Immune Strategy Of Complex Network

With the development of the Internet, the research on the virus propagation of complex networks has become an important research direction in complex networks. In Email network, social network and other complex networks, the scale of the network environment and people’s subjective behavior make the spread of the virus more complex. It promotes people to study of the spread of the virus characteristics, find out the main factors affecting the spread of the virus, and establish the model to portray the reality of life in the propagation of the virus. At the same time, in recent years, the researches show that the behavior of the virus propagation in network is closely related to the topology of the network, the small world and scale free characteristics of the real network affect the propagation behavior of the virus to a large extent. When the scale of the network increases, there is no positive critical value of propagation in scale free network. Even if the rate of virus propagation is very low, virus can be widely spread in the network. Therefore, it is of great practical significance to put forward effective strategies to control the outbreak and spread of the virus. Based on the research and analysis of the theory of complex network, this paper devotes to two aspects of the virus spreading model and immunization strategy. The main work is as follows:(1)We propose a multiple factors of email virus propagation model based on the detailed analysis of interactive email virus propagation model, combining with the mail virus characteristics and user habits, we introduce two concepts: the user’s response time and knowledge level. We define influencing factors’ parameters, design algorithm and program of email virus propagation simulation process. Through simulation experiments on Email networks and USAir97 network data set, we verify the rationality of our model, analyze the propagation process of the virus, and observe the changes of the number of infections under different parameters. At the same time, we also interactive email virus propagation model with the comparative experiment. Experimental results show that the improved model can effectively simulate the email virus propagation process, the user response time and knowledge level will affect the mail virus to spread to a large extent. Reducing the user response time, especially the nodes whose degree are large, and improving the user knowledge level can inhibit the propagation of email virus.(2)The study found that targeted immunization strategy can achieve better immunization but it is difficult to obtain the global information of the network is the limitation. Therefore, it is a good idea to get close to effect of the targeted immunization strategy by local information. Based on the detailed analysis of the AOC immunization strategy, we discuss random jump problem of entities in the search process, introduce the idea of simulated annealing, define the jump rules, and improve the immunization strategy based on AOC and simulated annealing. Through many experiments in real network data set, this paper found that the method can more efficiently and quickly find the nodes with large degree in network. We combine the interactive email virus propagation model and multiple factors email propagation model and compare the immunization efficiency of the two methods. By changing the proportion of the immunization nodes, we analyze the situation of node infection, and compare the immunization cost of the two methods. The experiments show that the improved method is better than the original method in the immunization efficiency, and verify the effectiveness of the proposed method.