河海大学学报(自然科学版)
河海大學學報(自然科學版)
하해대학학보(자연과학판)
JOURNAL OF HOHAI UNIVERSITY (NATURAL SCIENCES)
2013年
6期
554-559
,共6页
谭国平%霍孝义%谭林风%倪新洋%刘修泉
譚國平%霍孝義%譚林風%倪新洋%劉脩泉
담국평%곽효의%담림풍%예신양%류수천
稳健回归分析%迭代加权最小二乘法%预测模型
穩健迴歸分析%迭代加權最小二乘法%預測模型
은건회귀분석%질대가권최소이승법%예측모형
robust regression analysis%iteratively reweighted least-squares algorithm%prediction model
为探讨网络编码块长度、节点移动速度、接收节点个数对基于部分网络编码的实时多播( PNCRM)协议性能的影响,提出采用稳健回归分析构建预测PNCRM协议性能回归模型的方法。利用迭代加权最小二乘法求解回归系数,并对该模型进行显著性检验。结果表明,该模型对PNCRM协议性能预测的平均相对误差仅为4.83%,因而具有较准确的预测能力。
為探討網絡編碼塊長度、節點移動速度、接收節點箇數對基于部分網絡編碼的實時多播( PNCRM)協議性能的影響,提齣採用穩健迴歸分析構建預測PNCRM協議性能迴歸模型的方法。利用迭代加權最小二乘法求解迴歸繫數,併對該模型進行顯著性檢驗。結果錶明,該模型對PNCRM協議性能預測的平均相對誤差僅為4.83%,因而具有較準確的預測能力。
위탐토망락편마괴장도、절점이동속도、접수절점개수대기우부분망락편마적실시다파( PNCRM)협의성능적영향,제출채용은건회귀분석구건예측PNCRM협의성능회귀모형적방법。이용질대가권최소이승법구해회귀계수,병대해모형진행현저성검험。결과표명,해모형대PNCRM협의성능예측적평균상대오차부위4.83%,인이구유교준학적예측능력。
To investigate the influences of the BlockSize, the speed of node movement, and the number of receivers on the performance of partial network coding-based real-time multicast protocol ( PNCRM ) , a model for the prediction of the performance of PNCRM was built based on robust regression analysis. The iteratively reweighted least-squares algorithm was used to solve regression coefficients. A significance test was conducted for the model. The results show that the mean relative error was only 4.83% when using the model for the prediction of the performance of PNCRM, and, thus, the model has strong prediction capability.