Research Of Robustness Of Interdependent Networks

With the rapid development of science and technology, more and more complex network systems are derived, which makes our lives enter into the network era. Before 2010, the research on complex networks mainly focused on a single network. But now the modern infrastructures are coupled, interdependent and interacted by its complexed way. The most obvious characteristic of the infrastructures is that if one infrustructure is attacted, the coupled one will also produce corresponding fault. And then the cascading failure is produced in the whole coupling system. Based on this, people put forward the concept of interdependence network known as the second revolution of the complex network, and in the past 5 years they carried out extensive research. In 2010, Sergey V. Buldyrev and other scholors[11] proposed the theory model and mathematical analysis method of interdependent networks in Nature magazine, which has attracted wide attention of scholars at home and abroad. From then on, the interdependent network model has been constantly improved from one-to-one model to many-to-many model, from a global coupled model to partially coupled model, from the general to the cascade model load cascade model. At present, researches on the interdependent network mainly focus on robustness, cascade control and defense, attack strategy, the cascade model and the mathematical method of exploration. Among them, the study of the robustness of the interdependent network is the core.In this paper, we study the robustness of interdependent network mainly from its coupling and network structure. First of all, we explore the impact of coupling on the robustness of the network from the mathematical and simulation aspects, and put forward the probability model of dependence between the network and the network side connection. In the research of coupled scale free networks system, We found that: with ? decreases, percolation threshold cp also decreases; there is a threshold c?, if c? ??,the giant component is continuously approaching zero; ifc? ??,the giant component approaches zero discontinuously. But in the study of coupled scale free networks system, we found that compared with coupled scale free networks, parameter ? has little influence to percolation threshold. Secondly, we explore the effects of network structure on the robustness of the interdependent network, and put forward a interdependent network consists of multi-local-word network. We use the percolation theory and degree distribution theory as the theoretical analysis basis finding an interesting phenomenon: with the localization change of network structure, the percolation threshold had point mutation; Besides, there exists a critical region in which the network is in a safe state, but out of the critical region, the network suddenly turn into a vulnerable state.