基于FAR模型的网络通信量自适应预测

Adaptive Predictor of Network Traffic Based on FAR

本文分析了网络通信量中存在的自相似现象,并讨论了自相似通信量的预测及其对于网络性能分析的重要意义,提出了一种考虑了网络通信量自相似性的自适应预测方法,该预测方法在利用通信量自相似性的同时也考虑了通信量的短程相关特性,能够对实际通信量进行良好的预测。最后用实验仿真验证了这种方法在通信量预测中的有效性。

In this paper, the self-similar phenomenon in network traffic is analyzed. The prediction of self-similar traffic and its impact on the performance analysis of network system are also discussed- A adaptive predictor considered self-similarity of network traffic is proposed. This predictor takes account of the Long-Range Dependence and Short Range Dependence character of traffic at the same time, and can predict the real network traffic perfectly. At last some experiments confirm the validity of this predictor.