強化学習を用いた広帯域ミリ波最適スペクトル制御と先験知識を予測する技術の開発

利用强化学习和技术预测先验知识开发宽带毫米波最优频谱控制

• 批准号: 20J12528
• 负责人: 曹 誉文
• 金额: $ 1.09万
• 依托单位: Keio University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for JSPS Fellows
• 财政年份: 2020
• 资助国家: 日本
• 起止时间: 2020-04-24 至 2022-03-31
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-20J12528/
• 关键词: Deep learning; Resource allocation; Network control; Mobile edge computing; Analog beamforming; mmWave communications; deep learning; routing; caching; dense networks; federated learning

In this research-year, I focuse mainly on the following two research-topics: 1>. Low-overhead beam and power resource allocation using deep learning and its application in multiuser millimeter-wave (mmWave) communications; 2>. Deep learning-based network control and management in mobile edge computing (MEC) systems. Related research results have been published in high-quality journal and conference papers, respectively. Throughout our experiments, I generate images of resolution 4×4 and 8×8 and use these for distance estimation between users. Afterwards, I apply super resolution on images with size 4×4 to improve their resolution, and compare their results to the ones obtained with the original 8×8 images. For an area roughly equal to 60×30m, the proposed approach reaches an average mean squared error equal to 0.13 m.

在这个研究年度，我主要关注以下两个研究主题：1>。使用深度学习及其在多级毫米波（MMWave）通信中的应用及其应用的低空横梁和电源资源分配； 2>。移动边缘计算（MEC）系统中基于深度学习的网络控制和管理。相关研究结果分别发表在高质量期刊和会议论文中。在整个实验中，我生成了分辨率4×4和8×8的图像，并将其用于用户之间的距离估计。之后，我将超级分辨率应用于尺寸4×4的图像以提高分辨率，并将其结果与原始8×8图像获得的结果进行比较。对于大约等于60×30m的区域，提议的方法达到的平均平方误差等于0.13m。

项目成果

Spatial Degrees of Freedom Exploration and Analog Beamforming Designs for Signature Spatial Modulation

用于特征空间调制的空间自由度探索和模拟波束成形设计

• DOI: 10.1587/transcom.2020ebt0010
• 发表时间: 2021
• 期刊: IEICE Transactions on Communications
• 影响因子: 0.7
• 作者: CAO Yuwen;OHTSUKI Tomoaki
• 通讯作者: OHTSUKI Tomoaki

SUTD(シンガポール)

SUTD（新加坡）

Dual-Ascent Inspired Transmit Precoding for Evolving Multiple-Access Spatial Modulation

• DOI: 10.1109/tcomm.2020.3013030
• 发表时间: 2020-11-01
• 期刊: IEEE TRANSACTIONS ON COMMUNICATIONS
• 影响因子: 8.3
• 作者: Cao, Yuwen;Ohtsuki, Tomoaki;Quek, Tony Q. S.
• 通讯作者: Quek, Tony Q. S.

A Deep Learning-Based Low Overhead Beam Selection in mmWave Communications

• DOI: 10.1109/tvt.2021.3049380
• 发表时间: 2021-01-01
• 期刊: IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
• 影响因子: 6.8
• 作者: Echigo, Haruhi;Cao, Yuwen;Ohtsuki, Tomoaki
• 通讯作者: Ohtsuki, Tomoaki

Mobility-Aware Routing and Caching in Small Cell Networks Using Federated Learning

• DOI: 10.1109/tcomm.2023.3327278
• 发表时间: 2024-02-01
• 期刊: IEEE TRANSACTIONS ON COMMUNICATIONS
• 影响因子: 8.3
• 作者: Cao，Yuwen;Maghsudi，Setareh;Quek，Tony Q. S.
• 通讯作者: Quek，Tony Q. S.