Annual changes in topological structures of internet and its forecasts

Abstract

The internet is growing rapidly in recent years, during the growth of the internet, its topological structure is getting ever more complex. Because of the complexity of the internet, it is difficult to model. In order to help managing the internet, this project studies the annual change in the topological structure of internet. This knowledge is used to forecast the future growth of the internet. This project is a study about change of internet on AS level. The study is done by collecting topological data, computing topological metrics and analysing the relation between those metrics and time. The analysis is made up by two parts, degree analysis and spectral analysis. In degree analysis, only metrics related to degree are taken into account. For example, rank of node, frequency of node with certain degree. On the other side, spectral analysis mainly focuses on Eigen-values and Eigen-vectors of the adjacency matrix. From the study, Power Law Nature of Internet has been reconfirmed in the AS-Level. Also, by using the largest eigenvector of the adjacency matrix, a new existence of Power Law and an expanding core of Internet has also been discovered.