论文部分内容阅读
采用小波分析和神经网络工具对分时段网络流量进行预测,比基于顺序流量序列的预测方法具有更高的预测精度。首先将分时段网络流量序列进行小波分解后得到的各子序列分别用神经网络进行训练,然后将各子序列预测结果进行重构作为最终的预测结果。文章最后将不同的小波分解和分解水平的预测结果误差作了比较,指出应根据实际的网络流量序列的变化规律选择合适的小波;小波分解水平不宜过高,以避免重构误差的累加。
Using wavelet analysis and neural network tools to forecast network traffic in sub-periods, it has higher prediction accuracy than the prediction method based on sequential traffic sequences. Firstly, the sub-sequences obtained after the wavelet decomposition of the network traffic sequences in different time periods are respectively trained by the neural network, and then the prediction results of the sub-sequences are reconstructed as the final prediction result. At the end of this paper, we compared the different error of the wavelet decomposition and the prediction results, and pointed out that the appropriate wavelet should be selected according to the actual variation of the network traffic sequence. The wavelet decomposition should not be too high in order to avoid the accumulation of reconstruction errors.