基于神经网络的物联网连接数预测研究与实现  

A Research and Realization of number of loT connections Forecast Based on Neural Network

作　　者：文可鑫 李晋 周慧峰 祁钰 WEN Kexin;LI Jin;ZHOU Huifeng;QI Yu(China Mobile Group Design Institute Co.,Ltd.Chongqing Branch,Chongqing 400042,China)

机构地区：[1]中国移动通信集团设计院有限公司重庆分公司,重庆400042

出　　处：《长江信息通信》2024年第9期206-208,共3页Changjiang Information & Communications

摘　　要：随着信息技术发展,特别5G商用以后推动的物联网的迅猛发展,成为了下一个产业风口,促使生活、工作、学习等与信息通信的关系日渐紧密。物联网连接数的快速增长给通信运营商的网络建设带来了不确定性,文章构建了基于量化宏观因子融合神经网络算法预测模型,提升物联网连接数增长预测研究,提高通信运营商网络覆盖建设准确度。With the development of information technology,especially the rapid development of the Internct of Things promoted after the commercialization of 5G,it has become the next industry outlet,promoting the increasingly close relationship between life,work,study,and information and communication.The rapid growth of the number of Internet of Things conncctions has brought uncertainty to the network construction of communication operators.This article builds a prediction model based on quantitative macro-factor fusion neural network algorithms to improve the growth forecast of Internet of Things conncctions and increase the network covcrage of communication operators.Construction accuracy.

关 键 词：BPNN GAR 预测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]