摘要
随着5G的不断发展,万物互联时代即将来临。海量设备连接、海量业务请求、超高网络负载、复杂动态的网络环境等对5G系统优化提出了巨大的挑战。面对这些技术难点,人工智能(AI)算法表现了其独特的优势。首先对5G系统中基于深度学习的AI算法相比于传统算法的优势进行介绍;随后,针对多接入边缘计算和毫米波大规模多输入多输出(mmWave massive MIMO)系统中的AI算法应用进行详细的阐述,并对比分析了各种方法的优劣;最后,根据已有研究,总结了AI算法在5G实际场景中存在的不足,并对未来研究方向提出了展望。
With the continuous development of 5G,the era of the internet of everything is coming.Problems such as massive device connections,massive application requests,ultra-high network load and complex dynamic network environment pose great challenges to the optimization of 5G systems in the context of the internet of everything.Facing these challenges,artificial intelligence(AI)shows its unique advantages.Firstly,the advantages of deep learning driven AI algorithms in 5G system compared with conventional algorithms were briefly introduced.Then,the application of AI algorithms in multi-access edge computing(MEC)and mmWave massive multiple-input multiple-output(MIMO)system were described in detail,with advantages and disadvantages of each method being compared and analyzed.Finally,according to the existing research,the shortcomings of AI algorithms in 5G application scenarios were summarized and the future research directions were forecasted.
作者
章坚武
王路鑫
孙玲芬
章谦骅
单杭冠
ZHANG Jianwu;WANG Luxin;SUN Lingfen;ZHANG Qianye;SHAN Hangguan(Hangzhou Dianzi University,Hangzhou 310018,China;University of Plymouth,Plymouth PL48AA,UK;Research Center for Intelligent Network,Zhejiang Lab,Hangzhou 311121,China;Zhejiang University,Hangzhou 310007,China)
出处
《电信科学》
2021年第5期14-31,共18页
Telecommunications Science
基金
国家自然科学基金资助项目(No.U1866209,No.61772162)
英国皇家学会国际合作交流项目(No.IEC\NSFC\181300)。
