Attacked And Repaired To The Complex Network Invulnerability Strategy Research

In recent years, with the development of complex network researching, the studies of complex network in structural and dynamic have obtained many exciting results. Related researches show that the complex network has a strong resistance to random attack, but is extremely fragile to deliberate attack. The research of invulnerability on complex network become a hot point with more and more frequent attacks. It is an effective measure to improve the invulnerability with repairing the complex network after the attack to it.In this paper, we researched on the attack strategies, repair strategies and invulnerability of the complex network. We proposed a evolution model of the directed complex network and the attack and repair strategies according to its characteristics, then test the strategist on two networks:a generated network by the model and a real world network. Analysis results showed that the proposed strategies improve the attacked directed network invulnerability.The research content are as follows:1, Firstly, introduced the basic features and typical models of complex network, summarized its attack strategies and invulnerability optimization on its topology.2, Secondly, proposed a evolution model of the directed complex network and implemented generative process by programming. The increased arcs of the model are based on the in-degree and out-degree of nodes, the theory analysis and numerical simulation results proved the in-degree and out-degree distributions of the model are in line with scale-free network power-law distribution. Based on this, we studied the invulnerability when the model was subjected to the deliberate attack and was repaired.3, Thirdly, proposed an ideal of attack and repair based on the characteristics of complex networks, on the attack strategy, combined the removal methods with the in-degree and out-degree of nodes; on the repair strategy, in order to restore the connectivity of the nodes which were attacked, we proposed two kinds of repair strategies:random selection and preference selection, which are based on the betweenness of the current directed complex network and by changed the in-degree and out-degree of attacked nodes to one. Then illustrate the effectiveness of two repair strategies through the examples.4, Finally, we performed experiments on the attack and repair the strategies based on the generated network by the model and a real world network. The result shows that the proposed repair strategies can effectively improve the invulnerability of the attacked complex network.