改进ESNs在通信话务量预测上的应用研究  

Application research of improved ESNs in communication traffic prediction

作　　者：刘俊霞 刘智勇 刘文 LIU Jun-xia;LIU Zhi-yong;LIU Wen(School of Internet Finance and Information Engineering,Guangdong University of Finance,Guangzhou 510521,China;Department of Control Engineering,Xinjiang Institute of Technology,Urumqi 830023,China)

机构地区：[1]广东金融学院互联网金融与信息工程学院,广州510521 [2]新疆工程学院控制工程学院,乌鲁木齐830023

出　　处：《信息技术》2024年第5期38-45,51,共9页Information Technology

基　　金：新疆自然科学基金项目(2019D01A30);新疆高校科研计划青年自然科学基金项目(XJEDU2018Y056);国家自然科学基金地区科学基金项目(61962058)。

摘　　要：现有回声状态网络(Echo State Networks,ESNs)的通信话务量预测方法只考虑了历史通信话务量对预测性能的影响,较少涉及多个输入变量的通信话务量预测问题。文中首先针对ESNs用于实际多元时间序列预测任务时训练效率低,输入数据维数较多时计算复杂度大的问题,提出用改进的交替方向乘子算法(IAD-ESNs算法)训练ESNs;针对单一输入变量不能提供更加全面的预测信息,提出了改进ESNs的多变量预测模型(MP-IADMM-ESNs)。以真实通信话务量数据进行仿真实验,结果表明,提出的预测模型MP-IADMM-ESNs对多变量通信话务量预测有较高的预测精度和预测效率。Based on the fact that the existing traffic communication prediction methods for Echo State Networks(ESNs)only consider the impact of historical traffic on the prediction performance,and rarely involve the traffic prediction problem of multiple input variables.Therefore,this paper firstly proposes to use an improved alternating direction method of multipliers algorithm(IAD-ESNs algorithm)to train ESNs for solving the problems of low training efficiency and large computational complexity when ESNs are used in actual multivariate time series forecasting tasks.Secondly,as single input variable cannot provide more comprehensive prediction information,an improved multivariate prediction model for ESNs(MP-IADMM-ESNs)is proposed.Simulation experiments with real communication traffic data show that the proposed prediction model MP-IADMM-ESNs has higher prediction accuracy and prediction efficiency for multiple input variables.

关 键 词：多元时间序列 回声状态网络 时间序列预测 交替方向乘子算法 通信话务量 

分 类 号：TN915.0[电子电信—通信与信息系统]