Research On Important Node Identification And Robustness Of Complex Network Based On Topological Features

Nowadays,the normal running of society is increasingly dependent on the stable performance of various complex systems,including large-scale power grids,air and rail transportation networks,and the Internet,among others.In complex networks,critical node identification and robustness optimization are of great importance for disease prevention and control,information dissemination,disaster recovery,etc.,and have become important and popular topics in the field of complex networks and complex systems research.Classical node identification algorithms often require traversing all nodes to obtain global information about the network,which makes them difficult to apply to large networks;in addition,classical network optimization algorithms focus on strengthening the connected edges between core nodes in the network to optimize the network,but the optimized network does not perform well in the face of some deliberate attacks.Therefore,there is a need for the development of an effective method for the identification of important nodes and a method for the optimization of network robustness for larger networks.Based on the above analysis,this paper’s main works and innovations are:(1)Critical Node Identification Algorithm Based on Articulation Points and Betweenness Centrality APBCA new method for identifying important nodes(Articulation Points-Betweenness Centrality,APBC)is proposed by combining Articulation Points and Betweenness Centrality,which takes into account the network topology changes and uses different strategies in stages to maximize the selection of the most important nodes for maintaining network connectivity.By comparing with other classical identification methods,the simulation results show that the APBC method has the greatest damage to the network topology connection integrity after removing the selected nodes.In addition,the complexity of the algorithm to find the articulation points and calculate the betweenness centrality of them is much lower than that of calculating the betweenness centrality of all nodes.Therefore,APBC is a cost-effective method for identifying critical nodes that can be used in large networks.(2)Edge Rewiring Mechanism Based on Long-range Links and Articulation Points ALRAn edge rewiring method(Articulation Points and Long-range Links Rewrite,ALR)based on long-range links and articulation points is proposed,which focuses on strengthening the links between distant low-degree nodes while reducing the impact of attacks on articulation points in the network.The robustness of the optimized network against five different attacks is investigated by comparison with three other rewiring mechanisms.Simulation results show that the ALR method can effectively improve the robustness of the network against articulation point attacks compared with other methods,and the global network efficiency of the ALR-optimized network is improved the most,so the optimized network has faster global information transfer speed.