网络流量预测模型研究进展  

Research and development of network traffic prediction model

作　　者：邱婧[1] 夏靖波[1] 吴吉祥[1] 

机构地区：[1]空军工程大学电讯工程学院,陕西西安710077

出　　处：《计算机工程与设计》2012年第3期865-869,共5页Computer Engineering and Design

基　　金：陕西省自然科学基金项目(2009JM8001-1)

摘　　要：网络流量预测的关键是建立相应的数学模型。介绍了网络流量预测模型的基本概念和分类,指出了单一预测模型存在的问题。从模型构建原理出发,重点对组合预测模型进行了系统的分类,将其分为线性组合模型、优化组合模型、分解重构组合模型,介绍了各组合模型的相关研究进展,并分析了它们的特点和适用范围,指出组合预测模型能更精确、全面地预测流量。针对流量预测目前存在的问题,对未来研究趋势提出了几点思考。The key of network traffic prediction is to establish the corresponding mathematical model. Firstly, the basic concept and classification of network traffic prediction model is introduced, and the problems in the single prediction model are pointed out. Then, according to their different construction principles, the combination prediction model is systematically classified as the linear combination model, optimization combination model, decomposition and reconstruction combined model; Meanwhile, the related research progress of each class is introduced, their characteristics and the field of their application are analyzed. Final- ly, some suggestions are proposed for the future research directions of traffic prediction model accordhag to their current problems. Key words： traffic prediction; combination prediction model; linear combination model; optimization combination model; decomposition and reconstruction combined model

关 键 词：流量预测 组合预测模型 线性组合模型 优化组合模型 分解重构组合模型 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]