Network traffic classification method and device

Abstract

The embodiment of the invention provides a network flow type classification method. The method comprises the following steps: acquiring each to-be-classified network flow; inputting each to-be-classified network flow into a pre-trained target domain feature extractor; and inputting the target domain flow characteristics into a flow classifier so as to trigger the flow classifier to identify the types of the network flows to be classified according to the target domain flow characteristics and output the flow types of the network flows to be classified, thereby obtaining the target domain flowcharacteristics corresponding to the network flows to be classified. According to the embodiment of the invention, the pre-trained target domain feature extractor is used to extract the target domaintraffic features of the network flow to be classified. The flow classifier classifies the network traffic according to the flow characteristics of the target domain, and does not need to collect and mark the network flow with new flow characteristics in the training process, so that the manpower, material resources and time consumed are reduced.

Description

technical field [0001] The invention relates to the technical field of computer networks, in particular to a method and device for classifying network traffic. Background technique [0002] With the advent of the era of big data, new Internet applications emerge in an endless stream, and various types of network services are widely used in daily life. The amount of network business data has also increased sharply, so the composition of network traffic has become increasingly complex. In order to better implement network management and network security measures, network managers need to classify massive network traffic types. [0003] There are many existing network traffic classification methods, among which the classification method of network traffic type based on the neural network model is widely used in daily life. The classifier in the neural network model identifies the network traffic type. When the existing network traffic classification method uses the characteris...

AI Technical Summary

Problems solved by technology

With the wide application of the network, each type of network traffic will have new features with users at different times and places. For the new features of network traffic types, the existing network traffic classification methods need to collect network data with new features. Traffic is used as a training sample to perform classification training on the neural network classification model, which consumes a lot of manpower, material resources and time

Images