云计算下的一种网络流量预测算法的研究

Research into the Network Traffic Prediction Algorithm in Cloud Computing

为了提高云计算下的网络流量的预测精度,提出了一种基于遗传算法的小波BP神经网络的预测模型。首先针对BP神经网络的不足,引入动量项进行改进,其次,在遗传算法中加入进化操作,提高了算法的整体效率,将改进后的算法与小波函数进行融合,整体优化了BP神经网络的阀值和权值进行优化。仿真实验通过与其他文献算法的比较,说明本文算法具有良好的收敛速度,有效的提高预测的精度。

In order to improve the accuracy of predicting network traffic in cloud computing,a wavelet BP neutral network prediction mode based on genetic algorithm is proposed.First of all,aiming at the deficiency of BP neutral network,momentum is introduced for improvement.Secondly,evolutionary operation is introduced to genetic algorithm to improve the overall efficiency of the algorithm.The improved algorithm is integrated with wavelet function to optimize the threshold and weight of the overall BP neutral network.Through comparing algorithms in other literatures,the simulation experiment shows that algorithm in this paper has good convergence speed and can effectively improve efficiency of prediction.