基于智能优化的分布式网络流量预测方法

网络流量预测是网络管理的重要内容，高效的流量预测方法可提高网络管理效率。针对网络流量的时变性等问题，提出了一种基于智能优化的分布式网络流量预测方法。该方法采用果蝇算法优化3次指数平滑预测模型中的平滑因子，对时间窗口内收集到的网络流量进行预测，从而有效地提高3次指数平滑模型下网络流量预测的准确度与效率。仿真实验表明：相比传统3次指数平滑预测模型，此方法可解决平滑因子的不确定性所导致的预测结果误差问题，有效提高了网络流量预测精度。

Traffic Prediction Method Used in Distributed Network Based on Intelligent Optimization

Efficient network traffic prediction method can improve the efficiency of network management. On account of problems of network traffic if as burst, time-varying, nonlinear problems happen that caused by various coefficients, a distributed network traffic prediction method was proposed obeyed by in-telligent optimization. The fruit fly optimization algorithm was adopted in this method to optimize the smoothing coefficients of traditional triple exponential smoothing forecasting model. By predicting network traffic that is collected within time windows, this method effectively improves the efficiency of network traffic prediction. Simulation indicates that, compared with traditional triple exponential smoothing fore-casting model, the proposed prediction model can solve the problem of prediction error caused by smoot-hing coefficient. The optimal smoothing coefficient can be selected adaptively, thus improves the predic-tion accuracy.