基于改进Elmam神经网络的网络流量预测模型

Prediction Model of Network Traffic Based on Modified Elman Neural Network

提出了一种改进Elman动态回归神经网络,在此基础上建立了一种网络流量预测模型,该模型相对于传统的线性模型和BP神经网络模型具有更高的预测精度和更好的自适应性,利用某大学校园网统计得到的实际网络流量数据进行仿真实验,结果表明该模型具有良好的预测效果。

The modified Elman dynamic recurrent neural networks is proposed in this paper. Based on it, the network traffic prediction model is designed in this article. The model has higher precision and better adaptability than that of traditional linear model and BP neural network model. The perforanee of the proposed model is evaluted with the actual network traffic of an University schoolyard network. The experimental results show that the proposed model is effective and feasible.