基于深度学习的D2D毫米波通信中继选择
Relay Selection based on Deep learning for D2D Millimeter Wave Communication
投稿时间:2019-12-25  修订日期:2019-12-25
DOI:
中文关键词: D2D通信  中继选择  深度学习  CNN  能量效率
英文关键词: D2D  relay selection  deep learning  CNN  energy efficiency
基金项目:国家自然科学基金项目(面上项目,重点项目,重大项目)
作者单位E-mail
李中捷 中南民族大学 电子信息工程学院 lizhongjie@mail.scuec.edu.cn 
吴婉敏 中南民族大学 电子信息工程学院 wwm3505@163.com 
高伟 中南民族大学 电子信息工程学院  
摘要点击次数: 646
全文下载次数: 0
中文摘要:
      针对毫米波(mmWave,millimeter Wave)环境下,设备到设备(D2D,Device-to-Device)直接通信链路遇到阻碍时的中继选择问题,本文提出一种基于深度学习的方案.该方案采用卷积神经网络(CNN,Convolutional Neural Network)框架构建能够成功决策和精确预测的智能模型,并基于能量效率最大准则,实现最优化D2D中继链路选择.仿真实验采用文献[12]中毫米波多天线MIMO数据集DeepMIMO对本文所提模型进行训练和测试.仿真结果表明,本文提出的基于深度学习的中继选择方案在吞吐量和能量效率方面优于文献[4]中基于快速频率切换的传统方案.
英文摘要:
      Aiming at the problem of relay selection when the device-to-device (D2D) direct communication link encounters obstacles in the millimeter wave environment, this paper proposes a solution based on deep learning. This solution adopts a Convolutional Neural Network (CNN) framework to build an intelligent model, capable of successful decision-making and accurate prediction. Based on the maximum energy efficiency criterion, the most optimized D2D relay link selection is achieved. The simulation experiment uses the deep-MIMO dataset of millimeter wave multi-antenna MIMO in literature [12] to train and test the model proposed in this paper. Simulation results show that the relay selection scheme based on deep learning proposed in this paper is superior to the traditional scheme based on fast frequency switching in literature [10] in terms of throughput and energy efficiency.
View Fulltext   查看/发表评论  下载PDF阅读器
关闭