《电讯技术》，不收费还发稿酬的权威期刊、核心期刊！

首页 期刊视频编委会征稿启事出版道德声明审稿流程读者订阅论文查重联系我们English

引用本文：

张燕燕,王鹤鸣,姬天相,等.机器学习在5G超密集网络切换中的应用[J].电讯技术，2019，（12）： - . [点击复制]
ZHANG Yanyan,WANG Heming,JI Tianxiang,et al.Application of machine learning in 5G ultra-dense network handover[J].，2019，（12）： - . [点击复制]

本文已被：浏览 2238次下载 47次	码上扫一扫！
机器学习在5G超密集网络切换中的应用
张燕燕,王鹤鸣,姬天相,王军选
0 字体:加大+\|默认\|缩小-
(西安邮电大学通信与信息工程学院，西安 710121)

摘要:

超密集网络（Ultra-dense Network,UDN）作为5G网络架构的关键技术，其切换时延及无效切换等已成为网络发展的巨大挑战。提出了一种基于集成策略的机器学习算法，并结合用户的移动性数据，进行较高精确度的切换预测，减少切换时延及非必要切换等目标。仿真结果表明，采用结合改进机器学习算法的切换策略，不必要切换率降低了40.2%，平均时延降低了28.6%。

关键词: 5G 超密集网络；切换预测；机器学习

DOI：

基金项目:国家科技重大专项（2017ZX03001012-005）

Application of machine learning in 5G ultra-dense network handover

ZHANG Yanyan,WANG Heming,JI Tianxiang,WANG Junxuan

(School of Communication and Informatica Engineering，Xi′an University of Posts & Telecommunications,Xi′an 710121,China)

Abstract:

Ultra-dense network（UDN）is one of key technologies for 5G network architecture whose delay and invalid handover caused by handover have become huge challenges for network deployment.In order to reduce handover delay and unnecessary handover,an improved machine learning algorithm based on the user's mobility data is proposed to perform higher-precision prediction.The simulation results show that by using the handover strategy combined with the improved machine learning algorithm,unnecessary handover can be reduced 40.2%,average latency about 28.6%.

Key words: 5G ultra-dense network handover prediction machine learning