基于相空间重构和PSO-SVM的大型商场网络流量预测

Large Shopping Mall Network Traffic Prediction Based on Phase Space Reconstruction and PSO-SVM

准确高效的预测人流多变场景的移动网络流量,既可以保障网络安全稳定运行,又可以帮助运营商提前做好资源调度和分配。针对大型商场等人流多变场景移动网络流量预测问题,提出一种基于相空间重构的PSO-SVM预测方法。具体的,将网络流量数据进行预处理,通过相空间重构的方法构建学习样本,并采用支持向量机(SVM)对学习样本进行预测。同时,利用粒子群算法(PSO)、遗传算法(GA)和交叉验证优化支持向量机中的参数,得到SVM的最佳优化方案。实验结果表明,经相空间重构后的PSO-SVM模型具有较好的预测性能。

Accurate and efficient prediction of mobile network traffic in changeable scenes of human flow can not only ensure the stable operation of network security but also help operators to schedule and allocate resources in advance.Aiming at the problem of mobile network traffic prediction in large shopping malls and other scenes with variable traffic,a PSO-SVM prediction method based on phase space reconstruction is proposed.Specifically,the network traffic data is preprocessed,and the learning samples are constructed by the phase space reconstruction method,and the support vector machine(SVM)model is used to predict the learning samples.At the same time,particle swarm optimization(PSO),genetic algorithm(GA),and cross-validation are used to optimize the parameters in the support vector machine to obtain the best optimization scheme of SVM.The experimental results show that the PSO-SVM model reconstructed by phase space has good prediction performance.