Researches On Immunization Strategies In Complex Networks

Over a long period of time, infectious diseases are always the biggest threat to public health. About 13 million of people die owing to being infected by infectious diseases all over the world every year, especially in developing countries where half of the deaths is caused by infectious diseases. Nowadays, with the rapid development of economy and society, the communication among people becomes increasingly frequent, which would not only increase the probability of outbreak of infectious diseases globally, but also result in difficulties of which immunization strategy are used to prevent and control infectious disease. Meanwhile, the severe side effect caused by vaccination and the limited resource of society force us to minimize the number of vaccinated people. Therefore, proposing a better immunization strategy plays a pivotal role in precaution and control of infectious diseases.Recently, the development of complex networks provides researchers new perspective to seek and develop better immunization strategies. For instance, the degree centrality strategy based on degree heterogeneity can effectively control the spread of diseases by only implementing immune on Hub nodes. Based on previous studies, we have carried further research with regard to immune strategies in complex networks. Our main work is as follows:(1) Immunization strategy based on the critical node in percolation transition. We propose a novel and successful targeted immunization strategy based on percolation transition.Our strategy repeatedly looks for the critical nodes for immunizing. The critical node, which leads to the emergence of the giant connected component as the degree threshold increases,is determined when the maximal second-largest connected component disappears. To test the effectiveness of the proposed method, we conduct the experiments on several artificial networks and real-world networks. The results show that the proposed method outperforms the degree centrality strategy, the betweenness centrality strategy and the adaptive degree centrality strategy with 18% to 50% fewer immunized nodes for same amount of immunization.(2) Local immunization strategy based on the scores of nodes. We propose a successful immunization strategy only depending on local information. Differing from the traditionally local strategies that respectively consider the certain immunized node and its neighbors, the proposed strategy initializes the scores of each node with their corresponding degree values,and then recalculates the score of a certain immunized node based on the degrees of both itself and its nearest neighbors. After that, the certain immunized node tries to find a nonimmunized higher-score neighbor to replace itself. To investigate the efficiency of the proposed strategy, we conduct the experiments on several synthetic and real-world networks(including assortatively mixed and disassortative networks). We compare the proposed strategy(ULS and KLS) with the existing local immunization strategies(ULD and KLD) and also the degree centrality targeted strategy. For ULS, it has approximate efficiency compared to KLD in most of the network studied here and even outperforms KLD in high-average degree ER networks.KLS is less efficient than the degree centrality targeted strategy in sparse networks. However, KLS can ensure good performance in the majority of real-world and synthetic networks.To summarize, ULS holds the advantage of 0.00% to 19.54% compared to ULD and KLS has 8.11% to 49.99% improvement than KLD. KLS even outperforms the degree centrality targeted strategy in some networks.(3) A biologically inspired immunization strategy for network epidemiology. Well-known immunization strategies, based on degree centrality, betweenness centrality, or closeness centrality, either neglect the structural significance of a node or require global information about the network. We propose a biologically inspired immunization strategy that circumvents both of these problems by considering the number of links of a focal node and the way the neighbors are connected among themselves. The strategy thus measures the dependence of the neighbors on the focal node, identifying the ability of this node to spread the disease. Nodes with the highest ability in the network are the first to be immunized. To test the performance of our method, we conduct numerical simulations on several computer-generated and empirical networks, using the susceptible-infected-recovered(SIR) model. The results show that the proposed strategy largely outperforms the existing well-known strategies.(4) Inspired by the evolutionism of genetic algorithm, we put forward target immunization strategy based on genetic algorithm. We perform evolution operation on traditional target immunization strategies by utilizing the genetic algorithm that is modified by 6 types of mutation operators. The experimental results demonstrate that the immune effects of adaptive betweenness centrality strategy could also be further improved.