Network traffic classification system and method based on depth discrimination features

Abstract

The invention relates to a network traffic classification system and method based on depth discrimination features. The network traffic classification system comprises a preprocessing module and a model learning module, wherein the preprocessing module uses network flows with different lengths generated by different applications as input, and each network flow is expressed as a flow matrix with afixed size so as to meet the input format requirement of a convolutional neural network (CNN); and the model learning module trains the deep convolutional neural network by taking the flow matrix obtained by the preprocessing module as input under the supervision of a target function formed by a metric learning regularization item and a cross entropy loss item, so that the neural network can learnthe input flow matrix to obtain more discriminative feature representation, and the classification result is more accurate.

Description

technical field [0001] The invention relates to a network traffic classification system and method based on depth discriminant features, and belongs to the technical field of computer network and network traffic classification. Background technique [0002] As one of the basic technologies to enhance network controllability, network traffic classification plays a vital role in network supervision and network security. For example, network service providers analyze network traffic distribution by classifying network traffic, and then perform better QoS (Quality of Service) control; enterprise networks use traffic identification technology to control application access; Lawful Interception requires an understanding of the type of content being transmitted over its network in the first place. In terms of network security, network traffic classification is the core part of an intrusion detection system, which can detect abnormal traffic in the network, so as to take effective d...

AI Technical Summary

Problems solved by technology

DPI technology has the advantages of high recognition accuracy, but at the same time it faces the following disadvantages: (1) When encryption technology is applied to payload data, DPI technology will lose its effect; (2) DPI technology cannot identify unknown characteristic values, When the eigenvalues ​​of network applications change, the corresponding eigenvalue library must be updated in time; (3) Due to the need to check the payload content of each data packet, the time and space overhead of DPI technology is huge; (4) the data packet Load content inspection faces privacy protection issues

[0008] Existing network traffic classification methods based on deep learning technology often only use deep network models as feature extractors, which cannot effectively solve the problems of intra-class data diversity and inter-class data similarity in network traffic classification tasks, thus Difficult to classify network traffic more accurately

Images