Research Of Complex Network Topology Modeling And Topology Optimization

In the past few years, there has been increasing interest in studying Complex network as relevant to many areas of science. Through many verifying research and theoretical prove, the researchers in various areas are all founded that many complex network abstract from complicated systems, have surprising similarity in some properties. This thesis combines algorithm with computer, and use it into the complex network, and do some researches on the complex network topology. The main research objects are topology model and topology optimization. These two aspects have very important significance not only on the theoretic but also on the realistic application. Though the research on complex network topology structure, at one band we can know about and explain various network characteristics better, such as network congestion, virus transmission and network attack and so on. On the other hand, we can use the research results into the specific problems, for example is can design data communication network, which has better properties, and it also can make us have deeply understanding about the advantage and drawback of the existing network, in order to take effective to avoid occurrence of the risk. The content of this thesis is divided into two parts, introduced respectively as follows.The first part of this thesis research about the topology model of complex network, this part has analyzed the common characteristics that the complex network has appeared, and we have also made a simple analysis to the forming reason of these common characteristics. We derived that preferential attachment and dynamic growth are the common characteristics of the complex network topology, and then we have build the network model combine with Internet concretely, because firstly Internet it-self is a typical complex network, secondly it's relatively easy to obtain the statistic data of Internet, and conveniently to examine the experiment results. During this part, we have proposed the complex network topology model based-on message transmission, this model based on the two main characteristics of complex network, which are preferential attachment and dynamic growth. My creation is that the producing process of the model has simulates the grow process of the realistic network, and the network topology hassome extent hierarchical structure, the data on the statistic characteristics is almost alike with that has derived by actual observation. At last, according to the iterated algorithm of this model, there has realized a topology generator. The output network topology is compatible with the input format of NS2.The second part researches about the optimization of the complex network topology. Many complex networks in reality is generally been designed the optimization object in advance. These networks not only formed naturally, such as food link network of the ecosphere, but also formed by artificially, such as computer communication networks. By considered the cost and performance, specifically considered the link degree of the network and the mean shortest distance, I have studied the network topology optimization using genetic algorithm.