Network flow predicating system and flow predicating method based on neural network

Abstract

The invention discloses a network flow predicating system and flow predicating method based on a neural network, and belongs to the technical field of a computer. The network flow predicating system comprises a data collecting module, a data preprocessing module and a network flow predicating module, wherein a data collection sub module realizes real-time collection of various flow information in the network in a network mode based on port mirroring; the data preprocessing module respectively stores collected data and performs normalization processing on the collected data, so that the sample data value is between 0 and 1; pure data is provided for the predicating module. The flow predicating module determines the topological structure and the network parameter of a neural network for flow predication according to collected IP network flow data; a neural network method is used for predication; the predication result is obtained. The system and the method have the advantages that the monitoring detection and analysis can be performed on various backbone networks; network abnormal events in the network can be monitored and detected in real time; the advanced early warning on the network abnormal conditions is realized.

Description

technical field [0001] The invention relates to the technical field of computer networks, in particular to a neural network-based network traffic forecasting system and a traffic forecasting method thereof. Background technique [0002] With the rapid development of the Internet with IP as the core technology and the great enrichment of network services, the Internet has gradually developed into a global public information transmission platform that carries various services and serves various user groups. However, due to the inherent connectionless nature of the IP protocol and the best-effort (Best-Effort) service principle of the traditional IP network, the traditional Internet cannot provide users with effective quality of service (QoS) guarantees, nor can it achieve effective monitoring and monitoring of network resources. manage. Therefore, network monitoring and management has become a hot topic of research. [0003] To achieve network QoS control, it is necessary to...

AI Technical Summary

Problems solved by technology

However, since ARIMA is only suitable for dealing with linear wide stationary processes, most of the real network traffic does not meet the prerequisites of the ARIMA prediction model, so the prediction accuracy of the network traffic ARIMA prediction model is low, and it cannot accurately describe all the characteristics of the network

Images