Research on Efficient modeling of massive network traffic prediction based on Cloud Computing

Under the condition of large-scale interactive model, Cloud computing is easily to cause traffic congestion and interrupt data transmission in the network, the traffic is need to be predicted to prevent network congestion. Network traffic has the characteristics of time variation and non-stationary, Using traditional statistical analysis method to predict, it will produce serious distortion. A method of mass network traffic prediction based on quantitative recursive analysis in cloud computing model in this paper, the transmission link model of network traffic is analyzed, and statistical feature sampling method is used to collect the original flow information, phase space reconstruction of the collected traffic bit sequence streams, Extracting the correlation characteristic quantity of the flow bit sequence. In the cloud computing model, analyzing characteristics quantity of massive traffic rules in the high dimension space by quantitative recursive method, and accurate prediction of mass network is realized according to the regularity feature in quantitative recursive graph. Simulation results show that the proposed method can predict the internal traffic flow characteristics of network traffic accurately, the prediction accuracy of the network flow is higher, and the convergence is better.