首页 | 本学科首页   官方微博 | 高级检索  
     

基于随机神经网络的多步网络时延预测模型
引用本文:胡治国,张大陆,侯翠平,沈斌,朱安奇.基于随机神经网络的多步网络时延预测模型[J].计算机科学,2009,36(7):85-87.
作者姓名:胡治国  张大陆  侯翠平  沈斌  朱安奇
作者单位:1. 同济大学计算机科学与技术系,上海,201804;中国人民解放军65583部队,辽阳,111000
2. 同济大学计算机科学与技术系,上海,201804
基金项目:本课题受国家自然科学基金资助项目 
摘    要:网络时延的动态变化反映了网络路径的负载特征,对时延的精确预测是实施网络拥塞控制、路由选择的重要依据,建立了基于随机神经网络的时延预测模型,该模型克服了传统时间序列预测方法受随机干扰因素影响大、模型结构辨识过程繁琐,以及传统神经网络预测方法易于陷入局部极值、偏离全局最优的缺点.仿真实验表明,在提前单步和多步的预测中该模型比AR模型、RBF神经网络预测算法的准确度更高.

关 键 词:网络时延  RNN神经网络  预测
收稿时间:2008/10/14 0:00:00
修稿时间:2008/12/15 0:00:00

Multi-step Network Delay Prediction Model Based on RNN
HU Zhi-guo,ZHANG Da-lu,HOU Cui-ping,SHEN Bin,ZHU An-qi.Multi-step Network Delay Prediction Model Based on RNN[J].Computer Science,2009,36(7):85-87.
Authors:HU Zhi-guo  ZHANG Da-lu  HOU Cui-ping  SHEN Bin  ZHU An-qi
Affiliation:Department of Computer Science and Engineering;Tongji University;Shanghai 201804;China;The Chinese People's Liberation Army;65583 Troops;Liaoyang 111000;China
Abstract:Network delay is an important performance metric of IP network which reflects network path's workload characteristics.The precise prediction for network delay is an important basis on congestion control and route selection.A new multi-step prediction method was proposed for network delay prediction based on the random neural networks,this method overcomes the disadvantage traditional time-series method and neural network method.Compared with traditional RBF network and AR model,the experimental result indic...
Keywords:Network delay  Random neural network  Prediction  
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《计算机科学》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号