Traffic data classification method and device

Abstract

A traffic data classification method and device. The method comprises: collecting data packets, regrouping the data packets into a flow to generate traffic data, and according to service type marking which is conducted on some traffic data in the traffic data in advance, for each service category, forming a learning sample correspondingly, and setting the rest in the traffic data as a traffic data set to be classified; extracting a common numerical attribute feature set of each piece of traffic data in the traffic data set, and organizing the traffic data in the traffic data set into a flow record which is composed of the common numerical attribute feature set; and according to the learning sample, calculating a common numerical attribute feature set of each service category in the flow record by means of subspace clustering, and according to the common numerical attribute feature set of each service category which is obtained by calculation and the common numerical attribute feature sets of the traffic data in the flow record, conducting service type marking on the traffic data in the traffic data set. A new traffic service feature in a network is updated in time.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a flow data classification method and device. Background technique [0002] In the prior art, the service type classification of network traffic has a wide application range and extremely high practical value. It can accurately classify the network traffic data in the port with high bandwidth and high transmission rate in real time. Since different service types in the network have different requirements for network resources, and people have different management methods for network traffic of different service types, efficient and accurate network traffic service classification is the basis for operations such as network resource management and flow control. . [0003] The network traffic classification technology based on Deep Packet Inspection (DPI for short) needs to rely on the corresponding service type feature library, and the construction of the feature library itself...

AI Technical Summary

Problems solved by technology

[0004] In view of the problems that the DPI traffic classification technology in the prior art cannot update the new traffic service characteristics in the network in a timely manner, resulting in low update efficiency and easy degradation of accuracy, the present invention is proposed to provide a traffic data classification method and device

Images