Research On Topology Survivability And Transmission Of Complex Networks

By the development of information technology and the advance of social civilization, more and more large-scale networks emerge include Internet, peer-to-peer networks. As the networks are closely related to our daily life, it is of great significance to research the topology survivability and transmission of these networks. Recently, researching the large-scale networks based on complex network theory has become the inevitable trend. In contrast to simple and small networks as network models of previous researches, analytical approaches of complex network theory are different from the traditional networks analytical methods. It focuses on macroscopical phenomena result from the interactions of nodes and the interactions of individuals inside the system. With the in-depth research for complex networks, the investigation of complex networks has covered many areas and promoted development of different subjects.The ultimate goal of studying complex networks is to understand how topological properties affect the dynamical processes including packet transmission, game and so on. Topology survivability and transmission of complex networks are very important research issues. The thesis studies the topology survivability of complex networks and transmission of complex networks using methods of graph theory, statistical physics and computer simulation. The achievements in this thesis are as following:(1) An edge compensatory model based on the degree of nodesAlthough most of researches about topology survivability of complex networks have been proposed for creating a network that has optimal robustness, one is often instead faced with an existing network that cannot feasibly be redesigned, yet whose topology can be less modified by a given measure. The edge compensatory model reconstructs topology of scale-free network by adding new edges. Simulations show that the threshold of crash can increase due to the edge compensatory model when the nodes with the lowest degrees are connected with the new edges preferentially. Moreover, empirical results have demonstrated that hiding a small quantity of new edges contributes to the improvement of scale-free complex networks on attack tolerance.(2) A traffic model of two-layer complex networksPrevious works study traffic model of complex networks with a basic assumption that the network has a single layer structure. Many real-world complex systems can be described as two-layer or multi-layer structures. We introduce the idea of layers to establish a traffic model of two-layer complex networks. By comparing four different two-layer complex networks based on ER random network model and BA scale-free network model, we find that the physical layer is much more important to the network capacity of two-layer complex networks than the logical layer. Two-layer complex networks with a homogeneous physical topology are found to be more tolerant to congestion. Moreover, simulation results show that the heterogeneity of logical and physical topologies makes the packet-delivery process of two-layer networks more efficient in the free-flow state, without the occurrence of traffic congestion.(3) A static weight routing strategy based on the degree of nodesA static weight routing strategy based on the degree of nodes is proposed. The routing strategy assigns the weight of logical edges according to the degree of physical layer nodes and the strategy parameterβ, and then chooses the routing path with the minimum weight summation of logical edges. We can quantify the network capacity of different routing strategy on two-layer complex networks by the phase transition from free flow state to congestion state. We find that the optimum values ofβis associated with the topology of two-layer complex networks, whereas the network size is unrelated. Comparing with the traditional shortest path routing strategy, we find the routing strategy can greatly improve the network capacity and thus can reduce the congestion of two-layer complex networks.(4) A dynamic weighted routing strategy based on the queue lengthA dynamic weighted routing strategy based on the queue length is proposed. In the routing strategy, each logical edge is assigned a weight value that is related with the queue length of physical layer nodes and dynamic update the weight according to the variation of queue length,and then choose the routing path with the minimum weight summation of logical edges. Compared with the traditional shortest path routing strategy and the static weight routing strategy, the network capacity of two-layer complex network is further enhanced by the dynamic weighted routing strategy.