On Transmission Dynamics Of Complex Networks And Enhancing Of Load Capacity By Limited Resource

The small-world effect and the scale-free feature have been proposed to open the new chapter of complex network research. Many scholars in various fields are making research on the structure and dynamic behavior of complex network, due to many characteristics of real networks can well be reflected by complex network theory. Huge transmission networks such as communication networks, traffic network, and logistics network have played a more and more important role in human’s production and day-life. With the development of networks, which have the function of transmission, the pressure of load on networks is also growing. And it is more and more urgent that the transport on information or traffic should be keeping stably and efficiently. Therefore, the research on transmission behavior based on networks is one of the important problems of the dynamics on complex networks. In this dissertation, the basic theory of complex network and the transmission process of complex network are studied. Then, the critical value of phase transition in the transmission process is considered as the maximum load capacity, which is the most important index of the transmission performance. And it is analyzed the relationship between the network topology structure and the transmission performance. Then, it is researched how to improve the network transmission performance by mean of the reasonable allocation of resources with network resources limited. Meanwhile. several resource allocation schemes are put forward to improve the network load capacity and transmission performance.The main content and contributions of this dissertation are summarized as follows.1. To analyze the transmission dynamics process of complex networks. And it is been put forward that the research on the transmission problems of complex networks should base on analysis of betweenness.2. We set up a class of G-L routing strategies, which is using global information and local dynamic information of network. And the transmission performance of the network is significantly improved by the G-L routing strategy compared with the shortest path routing strategy. Then, the two flow models are constructed using the G-L routing strategies. According to the flow models, we comprehensively research on the network transmission load capacity in three groups of networks constructed by the ER random network model, the WS small-world network model, HK-BA scale-free network model.3. We research on the network transmission performance under the condition of the node resources limited. By constructing the node capacity resource allocation schemes: average allocation, allocation based on the degree, allocation based on the betweenness, we analysis the effect of three kinds of scheme for network performance, and find the network performance under the allocation scheme based on node degree is superior to the average allocation, and the allocation scheme based on betweenness is optimal scheme among the three kinds of allocation.4. We research on the network transmission performance under the condition of the edge resources (bandwidth) limited. We find that the network transmission performance renders accelerated decline, with the linear increase of bandwidth constraints in scale-free networks, when the bandwidth is limited. Then, we design the two kinds of heterogeneous random bandwidth allocation scheme; find that the classification random bandwidth allocation, with the reasonable combination, can effectively improve network load capacity.5. We propose a "counter-intuition" heterogeneous bandwidth allocation scheme. With an appropriate proportion of’controlled edge’, the bandwidth resource is reallocated in the network. According to the data flow model, with congestion awareness routing strategy, bandwidth allocation we proposed can adjust the data flow and improve the efficiency of bandwidth utilization. And we find that there is a strong correlation between the network traffic capacity and the bandwidth of vertex in the WS network.