The Study And Analysis Of Bittorrent Network System Model

In order to understand well the internal dynamics and related factors of the Internet, the establishment of an effective action model of BitTorrent Network users has great significance. The action model of BitTorrent Network users plays a vital role in the analysis and forecasting of the network performance, the designing network protocol, the simulation of the P2P Network and evaluating the performance, and the assurance of the quality of service.First, this paper records the changes in the number of new arriving users through the test of releasing several new torrent files to gather the data of user online time within 30 days. And then use the Poisson process model to fit the data of different time granularity in order to verify the regular scope of application of the new arriving users in the course of using Poisson process to describe BT Network. On this basis, the paper proposes the measure of using quadratic function to make the piecewise fitting of the data of new-arrival users within 24 hours and using the coefficient of determination to test the effect of the fitting so that to build the model of the laws of new-arrival users in BT Network.In addition, this paper established a mathematical model of BT Network to describe the influence on system performance caused by the ratio of the number of normal user nodes and attackers'nodes in the BT Network. Then according to this model, this paper analyzes the factors of deciding the users'income in the BT Network. And based on this mathematical model, the paper discussed the income data changes of the normal users caused by the changes of BT Network structure in the P2SP mode.Finally, the paper analyzes influence on the income data of normal users under three kinds of attacks in the BT Network in order to compare the differences of the influence on the normal users under these three types of attack patterns for the further step.