An Improved Scale Free Network Model And Its Application

In recent years, complex network as an important object to explore the complexity problem, has aroused people’s high interest in exploration, the structure of complex networks and topological characteristics become the focus of analysis. In real life, a lot of complex networks are scale free networks, and their degree distribution has the characteristics of thick tail. By using the scale free network, the real network has been constructed in a wide range of areas such as economy, biology and computer. Now according to the real network’s some internal features, In order to better simulate the real network more accurately, putting forward a variety of extension models of BA scale-free network to simulate the real network lays a foundation of the further research of inner characteristics and how to control the real network mechanism by the parameters. This article will mainly study it from three aspects:First of all, this paper introduces the background and significance of the research of complex network, the current research situation, and describes the main research direction of this topic. And briefly introduces the basic concepts of complex network theory, then focuses on several typical models of complex networks, and points out several classical evolution model of BA network.Secondly, according to the defects of basic BA network in practical applications, the evolution mechanism of the network model are improved. First proposed a growth scale free network model, then describes for the degree distribution of the numerical computing program, in these schemes selected an appropriate program to calculate the degree distribution of the network, and the network’s degree distribution was simulated with computer. Based on this model, an improved evolutionary model is proposed, which is based on the node threshold value, the intrinsic properties of nodes and the partial edge deletion, Making use of the MATLAB to simulate the evolution process of the traditional BA scale-free model and the improved BA scale-free model, compares and analyzes the simulation results of average path length and clustering coefficient of the two models. And this paper simulates and compares the reckless and fragility of the two models when facing random failure and deliberate attacks.Finally, this paper applies respectively the traditional BA scale-free model and the improved BA scale-free model to the Internet and then simulates the network performance based on the packet loss rate, throughput and delay rate, verifies the improved BA scale-free network for the feasibility and the accuracy of the Internet network simulation.