Research On The Life Characteristic Of Internet Macroscopic Topology

Internet is a complex network which is constructed by the humanity and the research on its topology evolution is the hot topic at present. Especially in recent years the achievement obtained in the complex network research aspect caused the people to make the very obvious progress to Internet topology. But these present research mainly concentrated on the analysis on the principle of dynamics and the behavior of synchronization among nodes, promulgated the intrinsic mechanism of its topology when its scale expanded. From macroscopic view, the evolution of Internet topology carried on not chaotically but orderlly. The evolution principle of Internet macroscopic topology was significant for constructing the module of Internet topology and other research on Internet.In this dissertation an evolution adaption model of the Internet macroscopic topology was established and the order parameter which affected the evolution of Internet macroscopic topology was solved. Based on the analysis on the order parameter it could be discovered that if the outside force was equal to zero or infinity, the Internet macroscopic topology was affected by the long-term survival effect. Internet evolved into a regular network; However if the outside force was equal to an unstable static topology evolution variant, Internet was affected by the short-term survival effect. Internet would evolve into a star network. At present Internet was affected by the long-term survival effect and short-term survival effect.Furthermore Internet macroscopic topology was studied in these three aspects -the information metabolism, the self-duplication and the catastrophy of Internet topology.The metabolism is one of the basic life characteristics. The Information metabolism of Internet maintained the existence of Internet topology by getting negative entropy flow from the outside. According to the theory of Shannon entropy, the Internet standard entropy was defined. By using the global Internet monitor data provided by CAIDA Skitter project, the changes of the Internet standard entropy with time were computed. The computing results using the monitor data get by Riesling showed that the Internet standard entropy reduced from the maximum 0.379 in April 2000 to 0.318 in May 2004. The reduction extent per month was 0.12 percent. On the other hand the computing results using the monitor data get by apan_jp showed that the Internet standard entropy reduced from the maximum 0.388 in April 2000 to 0.297 in May 2004. The reduction extent per month was 0.18 percent. Those showed that the Internet standard entropy tended to reduce, so the Information metabolism existed in the process of the evolution of Internet macroscopic topology.