Analysis of Cascading Failure Based on a New Load Redistribution Model

Cascading failure has drawn increasing attention to security problems due to the failure of a part triggers the failure of successive parts and even the entire network.In order to avoid cascading propagation more effectively, based on the local preferential redistribution rule of the failed node's load, researchers propose a new allocate strategy, considering the product of the degree and remaining capacity of the node's neighboring nodes.Here, researchers investigate the effect of the new allocate strategy for the robustness against cascading failures on a typical network, i.e.BA scale-free networks.Compared with two previous allocate methods, researchers find that our presented model is more robust than the previous models by the simulation and analysis in BA scale-free network.Moreover, researchers investigate how S attack changes depending on some of the parameters in our model.It is also found that the network robustness has a positive correlation with the value of α in the ranges of the larger avalanche, whereas, on the contrary, the smaller the value of α is the more robust the network is.Researchers can apply the result of the correlation between the parameters and the network robustness to allocate loads in real network.