A network traffic classification method capable of identifying unknown traffic and terminal equipment

Abstract

The present invention provides a network traffic classification method and terminal equipment capable of identifying unknown traffic. The method includes: when network traffic data passes through an initial network traffic classification model, judging whether the network traffic data is unknown traffic based on a threshold judgment method based on information entropy; unknown traffic Belongs to a new category; if the network traffic data is a known category of traffic, the dynamic adaptive algorithm based on the balance loss adds the ability to incrementally learn new categories to the initial network traffic classification model to obtain a real-time network classification model; the new category The loss of and the distillation loss of known categories are combined into a balanced loss as the loss function of the real-time network classification model. While accurately identifying known types of traffic, the network traffic classification model can also identify unknown types of traffic well. At the same time, it can dynamically adapt to the latest traffic in the current network environment, update itself in a timely manner, and always maintain a high classification accuracy , which is very suitable for real-time online network traffic classification tasks.

Description

technical field [0001] The invention relates to a network traffic classification method and terminal equipment, in particular to a network traffic classification method and terminal equipment capable of identifying unknown traffic. Background technique [0002] Internet and information technology is one of the fastest-growing technologies in the world of science and technology research. Since my country's access to the international Internet, in just over 20 years, with the unremitting efforts of various domestic Internet companies, China's Internet industry has gone from following the world to running in parallel with the world, and then to leading the world in the future. The great leap and the great achievements are obvious to all. As of June 2018, the number of netizens in my country has reached 802 million, and the Internet penetration rate has reached 57.7%, which is an astonishing speed of development. Internet technology is changing people's way of life because of i...

AI Technical Summary

Problems solved by technology

[0009] In order to solve the problem that the network traffic classification method in the prior art generally lacks the ability to identify new traffic of unknown categories and the ability to dynamically update itself according to the latest traffic, the present invention provides a network traffic classification method and a terminal that can identify unknown traffic equipment

Images