网络背景流量的分类与识别分析研究  

Classification and identification analysis of network background traffic

作　　者：易灿[1] Yi Can(Hunan Mass Media Vocational and Technical College,Changsha Hunan 410400)

机构地区：[1]湖南大众传媒职业技术学院,湖南长沙410100

出　　处：《电子测试》2020年第19期92-93,共2页Electronic Test

基　　金：湖南省自然科学基金课题“基于行为图谱的大规模web流量群体识别算法研究与应用(20J7015)”。

摘　　要：识别网络应用和分类相应流量的过程就是互联网流量分类,同时也是现代网络安全管理系统中最基本的。网络安全的基础技术就是流量分类,流量分类识别方法包括基于端口的预测方法和基于有效载荷的深度检测方法。文章从基于端口的识别分类和深度包检测的识别分类方面介绍了传统流量识别分类方法;进一步从数据及采集方法、有监督方法、半监督方法等方面分析了机器学习的识别分类。The process of identifying network applications and classifying the corresponding traffic is the classification of Internet traffic,which is also the most basic in the modern network security management system.The basic technology of network security is traffic classification.The identification methods of traffic classification include port-based prediction method and payloadbased depth detection method.This paper introduces the traditional traffic identification and classification methods based on port identification and deep packet detection.The identification and classification of machine learning are further analyzed from the aspects of data and collection method,supervised method and semi-supervised method.

关 键 词：网络背景 流量的分类 机器学习 基于行为模式的分类 

分 类 号：TP393.06[自动化与计算机技术—计算机应用技术]