本文選題：可用帶寬 + 在線預(yù)測��； 參考：《計算機應(yīng)用與軟件》2014年10期

【摘要】：針對目前端到端可用帶寬預(yù)測方面研究工作較少的現(xiàn)狀,提出一種基于核主成分分析KPCA(Kernel Principle Component Analysis)和最小二乘支持向量機LSSVM(Least Squares Support Vector Machine)的可用帶寬在線預(yù)測算法ABOP。在采集網(wǎng)絡(luò)狀態(tài)樣本數(shù)據(jù)并對其進(jìn)行相空間重構(gòu)的基礎(chǔ)上,采用KPCA對數(shù)據(jù)進(jìn)行降維降噪處理,最后基于LSSVM對可用帶寬進(jìn)行在線預(yù)測。為減小計算開銷,提出一種遞推計算的方法加快模型更新速度,并采用粒子群優(yōu)化算法對模型參數(shù)進(jìn)行多步更新,確保了在線預(yù)測的時效性。仿真表明,提出的ABOP算法具有較高的預(yù)測精度和較快的預(yù)測速度,能夠滿足可用帶寬在線預(yù)測的要求。
[Abstract]:In view of the lack of research on end-to-end available bandwidth prediction, an on-line prediction algorithm for available bandwidth based on kernel principal component analysis (KPCA(Kernel Principle Component Analysis) and least squares support vector machine (LSSVM(Least Squares Support Vector Machine) is proposed. On the basis of collecting the network state sample data and reconstructing the phase space, KPCA is used to reduce the dimension of the data and the available bandwidth is predicted online based on LSSVM. In order to reduce the computational overhead, a recursive computing method is proposed to accelerate the updating speed of the model, and the particle swarm optimization algorithm is used to update the parameters of the model in order to ensure the timeliness of on-line prediction. Simulation results show that the proposed ABOP algorithm has higher prediction accuracy and faster prediction speed, and can meet the requirements of on-line prediction of available bandwidth.
【作者單位】： 河南機電高等�？茖W(xué)校計算機科學(xué)與技術(shù)系;
【基金】：河南省教育廳科學(xué)技術(shù)研究重點項目(12A520019)
【分類號】：TP393.06

【參考文獻(xiàn)】

相關(guān)期刊論文 前4條

1 韋安明;王洪波;林宇;程時端;;IP網(wǎng)帶寬測量技術(shù)研究與進(jìn)展[J];電子學(xué)報;2006年07期

2 姜明;吳春明;張e，

本文編號：1841150