Network flow-predicting method and device based on wavelet package decomposition and fuzzy neural network

Abstract

The invention discloses network flow-predicting method and device based on wavelet package decomposition and a fuzzy neural network, wherein the method comprises the steps of: collecting historic measured data of network flow; decomposing the original network flow onto wavelets with different time scales by wavelet package conversion; reconstructing flow signals on all time scales so as to lead the data volume of the flow signals to be identical with that of original signals; by the fuzzy neural network, studying and predicting the flow signals decomposed onto all time scales; and summing thepredicted values of the flow signals on all time scales to obtain the predicted value of network flow.

Description

Technical field: [0001] The invention relates to an accurate network traffic prediction model, which is a network traffic prediction model based on the combination of wavelet packet decomposition and fuzzy neural network. This model is widely applicable to WAN backbone network and packet domain core routers in LAN, in order to provide basic suggestions for dynamic allocation of resources, network capacity planning, etc. It belongs to the field of computer network. Background technique [0002] Network traffic behavior prediction is an important research direction of network behavior, involving network capacity planning, network equipment design, network resource allocation and other aspects of network management, which is of self-evident importance. Therefore, the research on traffic forecasting methods and models has also attracted the attention of researchers, and various models have been proposed to describe and simulate traffic characteristics. [0003] The initial tel...

AI Technical Summary

Problems solved by technology

[0007] However, there are two main deficiencies in this hybrid model: on the one hand, the wavelet decomposition method only further decomposes the low-frequency part of the network traffic, and the high-frequency part is no longer analyzed in detail

The high-frequency part of the original traffic reflects the burst characteristics of the network traffic, and its severe jitter makes its prediction less than ideal, which in turn affects the prediction accuracy of the entire network traffic.

On the other hand, the signals decomposed into various time scales still exhibit non-stationary properties. Traditional time series forecasting methods are usually based on certain assumptions when predicting non-stationary time series (these assumptions may not be consistent with the real situation), which also makes There is a certain distance between the prediction accuracy and the expected value

Images