面向数据挖掘的网络流量分析及预测研究综述

Survey of Data Mining-oriented Network Traffic Analysis and Prediction Research

开展网络流量分析预测研究能够帮助网络运营商及时地进行网络运维调整及未来网络设计,合理规划网络流量调度,进而有助于构建良好的网络环境。近年来,数据挖掘技术被引入网络流量分析研究,学者从网络流量特性分析、网络流量分类、网络流量预测等多方面入手,提出了一系列的解决方案。针对基于数据挖掘的网络流量分析预测方法进行综述,归纳了数据挖掘技术在网络流量分类及预测方面的研究成果和性能评价指标,并对其特点及应用场景进行了对比,分析了现有方法存在的问题,并对网络流量分析预测未来的研究方向进行了展望。

Carrying out network traffic analysis and prediction research can help network operators to timely adjust network maintenance,design future network,and reasonably plan network traffic schedule,which will help build a network with good environment.In recent years,data mining technology has been introduced into network traffic analysis research.Scholars have proposed a series of solutions from various aspects such as network traffic characteristic analysis,network traffic classification,and network traffic prediction.This article summarizes the network traffic analysis and prediction methods based on data mining,sums up the research results and performance evaluation indicators of data mining technology in network traffic classification and prediction,compares its characteristics and application scenarios,and analyzes the problems of existing methods.The future research direction of network traffic analysis and prediction are prospected.