Mesoscopic Network Topology Structure And The Research Of Transport Efficiency

Transportation network is one of the application networks in the contemporary society and is closely related to our daily lives, such as transportations, water and electricity supplies, the biological gene regulatory network, energy transport of thermal material and so on. It is found that the mesoscopic topology structures of network are particularly important to network functions. We are concerned with how to improve network transport efficiency through tuning the mesoscopic topology structures.In this work, we studied the relations between network transport efficiency and the mesoscopic topology structures of networks in scale-free networks, random network. We found that the network transport efficiency increased with the cyclic coefficient in scale-free network while the transport efficiency is invariant. The result shows that we can improve the transport efficiency by means of the loop structures in transportation network, which has important insights for the practical new-generation transportation network.We have also investigated the statistical properties of U.S.A. airline networks and China airline network to study the relation between network transport and topology structure. Statistical properties, including the degree distribution, the shortest path, the cluster correlation degree, the synchronization of network, and the cyclic coefficient, are explored. We find that there are many similar properties between the two networks from the analysis. The degree distribution of them obeys the double power-low distribution, and their networks have a hierarchical structure. The coefficients of degree-correlation are negatively, r=-0.208for U.S.A. airline network and r=-0.47635for China airline network, which suggests that the air transportation network is disassortative. It is found that the transport efficiency is higher when the coefficient of assortativity is negative with larger cyclic coefficient. The loop structure would bring important influence for the network transport.In order to the further elucidate the relation between the loop structure and other topology parameters in network, we simulate scale-free network and random network and change the topology structures through XS algorithm, and then investigate the effects of the loop network structure on the shortest path. We find that the shortest path of network decreased when the loop structures increased in scale-free network, while the shortest path of network is invariant in the random network. The empirical networks, such as electronic mail network, the Internet and so on, are scale-free networks. We investigate the properties of topology network structures through seven groups of empirical data of Internet AS. The results reveal that Internet AS network is scale-free network and also a small world network. The shortest path of network is invariant when the cyclic coefficient of Internet AS network changes little. The loop structure could provide more paths to the network transport.Our conclusions provides insights into the improvement of network transport efficiency in the real networks, such as transportation networks, the Internet and so on, through tuning the mesoscopic loop structures.