Computer Engineering and Applications ›› 2014, Vol. 50 ›› Issue (13): 90-95.
Previous Articles Next Articles
WANG Xuesong1, ZHAO Yuelong2
Online:
Published:
王雪松1,赵跃龙2
Abstract: In order to obtain better predicting results of network traffic, a novel network traffic predicting model is proposed which uses the advantages of echo state network and autoregressive moving average. Firstly, the linear results of network traffic are obtained by autoregressive moving average while the nonlinear results of network traffic are obtained by echo state network, and then the predicting results of model are fused to obtain the predicting results of network traffic, finally, the simulation experiments are carried out by using some network traffic data. The results show that compared with other predicting models of network traffic, the proposed model not only has improved the predicting precision of the network traffic, and has good robust.
Key words: network traffic, autoregressive moving average, echo state network, predicting precision, error compensation
摘要: 为了获得更加理想的网络流预测结果,融合回声状态网络和自回归移动平均模型的优点,提出一种基于ARMA-RESN的网络流量预测模型。分别采用自回归移动平均和回声状态网络对网络流量线性变化特征和非线性变化特性进行建模与预测,对自回归移动平均和回声状态网络的预测结果进行融合,得到网络流量的最终预测结果,最后采用具体网络流量数据以及多个对比模型进行了仿真实验。仿真结果表明,相对于其他网络流量预测模型,ARMA-RESN不仅提高了网络流量的预测精度,而且具有更好的鲁棒性。
关键词: 网络流量, 自回归移动平均, 回声状态网络, 预测精度, 误差补偿
WANG Xuesong1, ZHAO Yuelong2. Network traffic predicting based on ARMA-RESN[J]. Computer Engineering and Applications, 2014, 50(13): 90-95.
王雪松1,赵跃龙2. 基于ARMA-RESN的网络流量预测[J]. 计算机工程与应用, 2014, 50(13): 90-95.
0 / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: http://cea.ceaj.org/EN/
http://cea.ceaj.org/EN/Y2014/V50/I13/90