Research On Multi-Focus Large Scale Network Simulation Abstract Technology Based On Node Merge

As the frequent outbreaks of large-scale security events, such as worms, the loss of Internet users grows fast. The macroeconomic early warning research of abnormal events in large-scale networks becomes more and more urgent. But now because of the huge influence and harm of security events, it is difficult to reproduce the actual event network. Simulating the network is one of the most important methods to study the network security events. While as the environment of the network getting more and more complexity and the need of network research getting more and more widely, how to improve the simulation efficiency and scale down the network has become a bottleneck.To simplify the simulation topology is an effective way to improve the simulation efficiency and simulation size. While to abstract network topology alone can not guarantee the facticity of network. Therefore, when reducing the scale of the network topology and the network simulation complexity, how to guarantee the facticity of network simulation has become a problem worthy to be researched.This paper briefly analyzes the network topology characteristics, traditional graph sampling algorithms and graph collapsing algorithms, then proposes a network topology simplicity method based on focus-zone with node merging. And brings the nodes-weight, links-weight to topology, combines the model of association degree based on focus-zone and the model of abstract degree based on focus-zone, and improve the Forest Fire sampling method in order to ensure the key structure of original topology; After sampling, combining with the application features of security incidents, the nodes deleted merge into a "super node" attached to the upper router. Then through sampling network topology experiments and UDP worm simulation experiments, it can be shown that this method can reduce the scale of the network to some extent, greatly save a lot of resource overhead, improve the simulation efficiency and guarantee the facticity of the simulation.