Analysis And Modeling Of Mutation And Gentle Evolution Based On IPv6Topology Features

The evolution of the Internet macroscopic topology is an important part of the complex network theory. It has been widespread focused by scholars and has made considerable development in this area. To grasp the evolution of the macro-topology helps understand the regular of the formation and establish the model that according with the real network. Accompanied by the depletion of Ipv4address, Ipv6is about to usher in a period of vigorous development. Study of the evolution of modeling for Ipv6IP level topology has just started and the result of the research will have substantive significance on the future deployment of Ipv6networks.A comprehensive understanding of the topology evolution is necessary to establish the topology model that can simulate the real network. The evolution trend of IP level topology is gentle in most of the time. Therefore, the current study only considered the gentle evolution of topology with time series and ignored the jump changes. This paper introduces the outlier detection method and defined the jump changes in the topology evolution as mutation.This paper statistics data of topology evolution on the time span based on the authority Ipv6IP-level data by CAIDA(The Cooperative Association for Internet Data Analysis),and extract gentle and mutation evolution trend. The mutation characteristic was analyzed from the perspective of fine-grained topology. The conclusion was drawn by analysis:The mutation of the shortest path between September2011and October was caused by the "long side" and node number. Besides, we also find that the mutation of Internet topology is closely related to the social events and made a deduction that the IP-level topology happened easily to mutate. The proof was given by the Ipv6IP-level topology data between Egyptian Revolution and the earthquake in Japan.To simulate the real evolution on the basis of the study of mutation evolution, this paper establish IP-level topology evolution model—PLM (Power Law Model) to resolve the problem that the current model can’t simulate the mutation evolution of Internet topology. Contrast to static network generated by other models verify that the static validity of the PLM model and the PLM model network closer to the true Ipv6IP level topology. In terms of dynamic effectiveness of PLM, first according to the evolutionary process of the real Ipv6IP-level topology data to simulate the gentle trend through add factors that die and new in the PLM model. Based on the study on mutation analysis, some point was triggered and the phenomenon of mutation was simulated. The simulation results show that the PLM model can simulate the gentle evolution and part of mutation trend of the real Ipv6network topology evolution.