基于支持向量机的网络流量建模及预测

流量建模与预测对于大规模网络的规划设计和网络资源管理等方面都具有积极的意义,是网络流量工程重要组成部分。该文结合网络流量的时间序列特性,提出一种基于支持向量机的网络流量预测算法,并针对由于支持向量机采用先验知识选择参数会导致不同数据对先验知识适应程度不同,给出了一个动态调整优化参数策略。实验仿真结果表明,该算法具有很好的预测精度和适用性。

Network Traffic Model and Prediction Based on Support Vector Machine

Network traffic model and prediction is significant for design of network and resource management, and it is an important part of traffic engineering. In this paper, a method of traffic prediction based on support vector machine is presented with characteristic of time series of network traffic, and strategy of dynamic optimizing parameter is given because of the defect of parameter selection based on prior knowledge. The experiment simulation result s show that this method is of high precision and applicability.