Multi-neural network-based traffic matrix estimation method

Abstract

The invention provides a multi-neural network-based traffic matrix estimation method, which can improve the accuracy of the conventional network traffic matrix estimation. According to the method, the problem of memory fading or deformation of the traditional neural network used in traffic matrix estimation can be effectively overcome by respectively training sampled traffics before classifying. The errors of the multi-neural network-based estimation is remarkably less than that of estimation based on the traditional neural network.

Description

technical field [0001] The invention relates to the field of network measurement and neural network, in particular to a flow matrix estimation method based on multiple neural networks. Background technique [0002] The traffic matrix is ​​an overview of the traffic of the whole network. The elements in the matrix represent the traffic (OD traffic) starting from one node (source node) and terminating at another node (destination node) in the network. This source and destination node pair is also called an OD pair. The measurement and calculation of OD traffic is of great significance in the research and engineering practice of network structure configuration, management, and network traffic engineering. The research on the measurement and calculation of OD flow has received extensive attention from the theoretical and industrial circles at home and abroad. Since the traffic matrix needs to capture the global state of network traffic, direct monitoring is very expensive and ...

AI Technical Summary

Problems solved by technology

These will inevitably lead to a very slow training speed of the neural network, and the training of a large number of high-dimensional sample data will exacerbate the contradiction between the memory and plasticity of the traditional neural network, that is, the problem of memory distortion or memory disappearance

Images