摘要
为了更加精确地对网络流进行预测,降低网络拥塞的频率,提出了极端学习机的网络流量预测模型。针对网络流量混沌性分别确定原始网络流量的延迟时间和嵌入维数,采用极端学习机对网络流量的变化特点进行拟合,并对标准学习机进行改进,改善学习速度和预测性能,通过网络流量数据的预测实验验证其可行性。结果表明,与其它网络流量预测模型相比,改进极限学习机的网络流量预测结果更加可靠,对网络流量将来变化趋势可以更加准确描述,提高了网络流量预测精度。
In order to obtain an accurate prediction of network traffic flow and reduce the congestion frequency of network,a novel network traffic prediction model based on improved extreme learning machine is proposed in this paper.Firstly,the delay time and embedding dimension are determined according to the chaos of network traffic,secondly,extreme learning machine is used to simulate the change rule of network traffic where the standard learning machine is improved to increase the learning speed and performance,finally,the feasibility is verified by the network traffic data.The results show,the network traffic prediction results of the proposed model are more reliable compared with other prediction models,and can describe the change trend of network traffic and improve the prediction accuracy of network traffic.
作者
鲁华栋
时磊
Lu Huadong;Shi Lei(Network Management Center, Henan Polytechnic Institute, Nanyang 473000;Department of Electronic and Information engineering, Henan Polytechnic Institute, Nanyang 473000)
出处
《微型电脑应用》
2018年第4期53-56,共4页
Microcomputer Applications
关键词
网络流量
相空间重构
极端学习机
混沌变化特性
Network traffic
Phase space reconstruction
Extreme learning machine
Chaos variation characteristics
