Network traffic identification model based on deep learning in high-speed network environment

Abstract

The invention belongs to the technical field of network flow classification and specifically relates to a network traffic identification model based on deep learning in a high-speed network environment. The system comprises a switch, a controller, a flow acquisition module, a network flow feature extraction selection module, a network flow identification module, a deep learning module, a trainingsample library and an output module. The switch is in signal connection with a controller, wherein the controller is electrically connected with a flow acquisition module, the flow acquisition moduleis in signal connection with a network flow feature extraction selection module, the network flow feature extraction selection module is electrically connected with a network flow identification module, and the network flow identification module is electrically connected with a deep learning module and a training sample library. The network flow recognition model based on deep learning in the high-speed network environment is better in classification effect, capable of effectively dividing network flow in time and achieving fine-grained management and control, higher in learning capacity and higher in processing speed.

Description

technical field [0001] The invention relates to the technical field of network traffic classification, in particular to a network traffic identification model based on deep learning in a high-speed network environment. Background technique [0002] Network traffic is the amount of data transferred over a network. The size of the network traffic is of great significance to the design of the network architecture. Just as the width and connection mode of the road are designed according to the number and flow direction of the vehicles, it is very necessary to design the network according to the network traffic. Network traffic classification has been an important branch of the Internet in the past few years. Real-time network traffic classification is of great significance for network operators to optimize operations and manage networks. However, the existing classification methods often have poor classification results and relatively slow classification speed Slow, weak learni...

AI Technical Summary

Problems solved by technology

Network traffic classification has been an important branch of the Internet in the past few years. Real-time network traffic classification is of great significance for network operators to optimize operations and manage networks. However, the existing classification methods often have poor classification results and relatively slow classification speed Slow, weak learning ability, slow processing speed

Images