Load Redistribution Strategy Based on Dynamic Information to Cascading Failures

The global collapse caused by overload is a hot topic in the research of the robustness of thedependent networks. Compared with the homogeneous dependent networks, the research of the robustnessof the heterogeneous dependent networks is less. In this paper, based on the study of the robustness of thehomogeneous dependent networks, the dependent networks composed of different types of heterogeneousnetworks are generated, and the corresponding strategies are proposed to improve the network robustness inthe case of overload. Simulation tests show that the overall robustness of the dependent heterogeneousnetwork is affected by the degree of heterogeneity of a single network, and highly related to the types ofconnections between networks. When the connection between networks is function dependent, theconnection strength should be reduced in case of cascading failure; the load redistribution strategy based onthe medium number is easy to cause the failure of the central node and reduce the overall robustness of thenetwork. Simulation results show that the strategy is effective to improve the overall robustness of thedependent heterogeneous network