Research On Complex Of Internet Topology

The research on complex networks increasely develops recently. Especially after proposed the modeling of small-world and scale-free networks, more and more researchs focus on the complex networks. The small-world networks have not only the same high clustering coefficient as the normal networks, but also the same small average path length as the random networks. The degree distribution of scale-free networks presents the power law and has not the particular scale characteristic. Meanwhile, it also can explain the "rich gets richer" phenomenon observed in many real-life complex networks. The statistic characteristics of many real-life complex networks already discussed, economy networks, world-wide airport networks and scientific collaboration networks and so on. The statistic characteristics indicate that these real-life networks have both small world and scale-free features.Internet the important real-life complex networks is the hot topic in research of complex networks, especially research on complex of Internet topology. There are many statistic characteristics and models welling up in this area, which provide the scientific bases of using and managing the Internet.In this paper, first, we introduce the main models and those of statistic characteristics of complex networks including small world networks and scale free networks. Second, we introduce the main results of researchs of Internet topology. Then for most networks in engineering and technology, the interplay of system dynamics, data traffics and network topology is critical to the network evolution and performance evaluation. A new model is proposed for describing the growth of local-world weighted complex networks. This model combines the new vertices and new edges with the dynamical evolution of the weights locally, thereby generating a growing network with many statistical properties observed from real-world network examples. In particular, the model yields non-trivial time evolution of various vertex properties, including such as exponential and scale-free distributions of weights, strengths and degrees. Finally, we compare the clustering coefficient and correlation of router level Internet topology with those of our model resulting in validating the local evolving feature in the process of Internet topology growing.