Research Of Complex Network Modeling

Complex networks has become a hotspot of research in the past two years, he can describe most systems which have existed. Two kinds of the most typical characteristics of complex network is a small world phenomenon and scale-free properties. Combing with statistical method and data mining method to study the complex network modeling.The key research object of this paper is complex network topology modeling and topology optimization which are important in the research of theoretical and practical application.Complex network is composed of a variety of systems and the realiety complex systems is composed of thousands of unit interaction ralation.So that we will abstract unit as the vertex in the network and abstract relationship between units as the edge of the network.According to the study of network structure which is abstracted we can analyze the functions and properties of the original system.In recent years,reserchers also also made significant progress in these directions.This article uses the method of tatistical analysis to analyze and establish new network model to simulate the real network data which can help to predict the spread of the internet dynamic phenomena.Specific research work are as follows:Analyzing the small world phenomenon of the complex network and scale-free features deeply.Small world phenomenon is one of the most effective information transfer mode to reveals the complex network of sports objectively. Generally,if the network clustering coefficient is much larger than the corresponding random networks and the average paths nearly the same so that this network has small world phenomenon.An important index of small world phenomenon is clustering coefficient and average path length.For the scale-free network,if the degree of the node satisfy power-law distribution,then we will recognize this network is scale-free networks.Power-law distribution is different from poisson distribution and gaussian distribution.The tail of the former decrease slowly but the latter exponential decline.We can infer from above that the network center node will appear.Based on the research of the scale-free network model, and puts forward some views to the network modeling from a new Angle.A hypothesis was put forward,and has been demonstrated on some level.