| 摘要: |
| 针对高移动用户在毫米波系统中用户与基站之间频繁切换增大延迟开销问题，提出了一种新颖的集成深度学习协调波束成形（Deep Learning Coordinated Beamforming,DLCBF）解决方案。该方案通过协调多个基站同时为一个移动用户服务，接收用户上行导频序列和预编码码本训练模型，进而预测最佳下行波束向量，克服了大型天线阵列毫米波系统中选取最佳波束成形向量需要的巨大训练开销。仿真结果表明，所提方案与传统的毫米波波束成形策略相比有更高的频谱有效率。 |
| 关键词: 毫米波通信 波束成形 深度学习 |
| DOI： |
|
| 基金项目:国家科技重大专项(2017ZX03001004) |
|
| Beamforming of Millimeter Wave System Based on Deep Learning |
| LONG Ken,WANG Yaling,CHEN Xing,WANG Yi,TAN Luyao |
| (School of Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China) |
| Abstract: |
| A novel integrated deep learning coordinated beamforming(DLCBF) solution is proposed to address the problem of frequent switching between users and base stations in millimeter wave systems with high mobile users,which increases the latency overhead.By coordinating multiple base stations to serve a mobile user and receiving the user uplink conduction sequences and pre-coded codebook training model,the DLCBF solution can predict the optimal downlink beamforming vector,thus overcoming the large training overhead required to select the optimal beamforming vector for large antenna array millimeter wave systems.Simulation results show that the proposed solution is more spectrally efficient than conventional millimeter wave beamforming strategies. |
| Key words: millimeter wave communication beamforming deep learning |
《电讯技术》编辑部 