Research On Immune Strategies For Complex Networks

Immunization strategy,as one of the methods to evaluate the important of the individuals in complex networks,plays an extremely important role in the selection of immune objects to achieve faster containment of the spread of virus.However,due to the different networks and propagation characteristics in complex networks,the practicability of immunization strategy is limited and it is difficult to apply to different scenarios.Community networks are generated based on the clustering of complex networks,which are characterized by the aggregation of individuals with the same features.However,most of the previous studies considered immunization strategies with homogeneity within the community,but the actual networks are not.In addition,factors such as failure of information transmission between networks and interference of signals connection,which results in uncertain networks.Uncertain networks are common type of networks in real life,but they have never been involved in virus transmission scenarios,leading to the lack of immunization strategies in inhibiting virus diffusion.Therefore,the core of this paper is to design appropriate immunization strategies for different networks and propagation scenarios.On the basis of summarizing the existing research works,this paper proposes corresponding immunization strategies based on the centrality measurement and the direct relationship between the largest eigenvalue and the epidemic threshold for the community networks and uncertain networks.The change of the scale of infection under different immunization strategies are analyzed separately.The main content include:1)Overlapping community is a common type of community in social networks,which is characterized by the fact that an individual can belong to multiple communities at the same time.In this paper,overlapping nodes are used as the edges of the inter-community connection,and the community network is processed and reconstructed.The matrix model is transformed according to the proportion of overlapping nodes in community and the immunization strategy is designed according to the centrality measurement.This paper guarantees the reliability of immune objects by verifying the dual importance of the community in the network and the nodes in the community.The suppression of immunization strategy on the scale of infection is verified by method comparison.2)In the case where the network structure is determined,the epidemic threshold? is closely related to the largest eigenvalue ?,that is,the smaller the ?,the more difficult the virus is transmitted.In order to study the suppression of virus spread by immunization strategies under uncertain networks,this paper first proves that the relationship between ? and ? is still valid under uncertain networks through experiments.By minimizing ?,this paper combines the characteristics of eigenvector,node degree and the density of the network to design the immunization strategy.According to different uncertain edges,sampling method and extracting representative instance are used to verify the scale of infection under small-scale networks and large-scale networks.