| As the basis of Internet development and exploitation on higher levels, the Internet topology modeling can be used to analyze the Internet's large-scale topology, and forecast or control network's behaviors.The Internet is a scale-free, small-world network, which have a loose hierarchical structure. The Internet topology modeling starts from the random model to the hierarchical model, then to the evolution model. After researched deeply many current classical topology models, this paper finds that anyone can only describe parts of the networks' topology character.A new Internet topology model -Clique-VGBA model- is presented which ameliorate the current GBA model three times. Firstly, the new model takes into account of adding edges and deleting edges in the evolving networks to amend GBA model. Secondly, the idea of "preferential cut" is introduced to new model, because deleting edges more happens to nodes with less degree. Finally, new model combines the hierarchical model and the evolution model to accomplish Clique-VGBA model.Many characteristics of topology are analyzed with the corresponding metrics, including Clique-VGBA model's power law exponent, shortest path, clustering coefficient. Then these metrics are validated in realistic data of the Internet, and compared with the ER model, BA model, and GBA model. The analysis shows the new model can describe the Internet's topology character more maturely than the other models.The IM network is a virtual and social network in the Internet, which not only shows the Internet's statistical character, but also is a social network. This paper applies the Clique-VGBA model to the IM network, and analyzes its peculiar metrics, such as the betweenness, in-degree distribution, out-degree distribution, and so on. Furthermore, the paper analyzes the security of the IM network and presents some effectual security strategies. |
