AI based optimization of the spectrum and energy efficiency for Intelligent 6G

基于 AI 的智能 6G 频谱和能源效率优化

• 批准号: 22K14263
• 负责人: 李 傲寒
• 金额: $ 2.83万
• 依托单位: The University of Electro-Communications
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Early-Career Scientists
• 财政年份: 2022
• 资助国家: 日本
• 起止时间: 2022-04-01 至 2025-03-31
• 项目状态: 未结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-22K14263/
• 关键词: Intelligent 6G; AI based optimization; Spectrum efficiency; Optimization; Spectrum Efficiency; Energy Efficiency

During the past year, I have mainly conducted research on two topics. That is:1) Joint dynamic spectrum access and multiplexing techniques for optimizing the spectrum and energy efficiency.2) Automatic communication parameter decision based on AI in a practical uncertain wireless environment.For the first topic, a quantum annealing-based resource allocation was proposed for dynamic non-orthogonal multiple access (NOMA) systems. Specifically, an optimization objective problem is first formulated, considering the dynamic spectrum access and NOMA. Then, the optimal transmission parameters, including channel and transmission power, were derived through an exhaustive search and quantum annealing methods. Compared with the spectrum efficiency while jointly considering the dynamic spectrum access and NOMA to that without a joint consideration, it is clarified that the spectrum efficiency can be improved.For the second topic, several AI algorithms were proposed to determine the transmission parameters in a decentralized manner, including the channel, power, and spreading factor. The AI algorithms include laser chaos-based multi-armed bandit, Tug of War dynamics, and deep reinforcement learning methods. In addition, the proposed algorithms were verified by simulation and experiments conducted in a practical uncertain wireless environment.Based on the research results, 1 journal paper and more than 3 related international papers have been published. In addition, 2 journal papers have been submitted.

在过去的一年里，我主要就两个主题进行了研究。针对第一个问题，提出了一种基于量子退火法的动态非正交多址(NOMA)系统资源分配方法。具体地说，首先建立了考虑动态频谱接入和NOMA的优化目标问题。然后，通过穷举搜索和量子退火法，得到了包括信道和发射功率在内的最优传输参数。与联合考虑动态频谱接入和NOMA时的频谱效率进行了比较，说明频谱效率是可以提高的。对于第二个问题，提出了几种分散确定传输参数的人工智能算法，包括信道、功率和扩频因子。人工智能算法包括基于激光混沌的多臂强盗、战争动力学拔河和深度强化学习方法。此外，通过在实际的不确定无线环境中进行的仿真和实验对所提出的算法进行了验证，在此基础上发表了1篇期刊论文和3篇相关的国际论文。此外，还提交了2篇期刊论文。

项目成果

Uplink Grant-Free NOMA Using Laser Chaos Decision Maker

使用激光混沌决策器的上行链路无赠款 NOMA

• 发表时间: 2022
• 作者: Hasegawa So;Kitagawa Ryoma;Li Aohan;Kim Song-Ju;Watanabe Yoshito;Shoji Yozo;Hasegawa Mikio;Aohan Li;Aohan Li
• 通讯作者: Aohan Li

IEEE VTC2022-Fall (IEEE 96th Vehicular Technology Conference)

IEEE VTC2022-秋季（IEEE 第 96 届车辆技术会议）

• 发表时间: 2022

Deep Reinforcement Learning Based Resource Allocation for LoRaWAN

LoRaWAN 基于深度强化学习的资源分配

• 发表时间: 2022
• 作者: Hasegawa So;Kitagawa Ryoma;Li Aohan;Kim Song-Ju;Watanabe Yoshito;Shoji Yozo;Hasegawa Mikio;Aohan Li
• 通讯作者: Aohan Li

Multi-Armed-Bandit Based Channel Selection Algorithm for Massive Heterogeneous Internet of Things Networks

• DOI: 10.3390/app12157424
• 发表时间: 2022-08-01
• 期刊: APPLIED SCIENCES-BASEL
• 影响因子: 2.7
• 作者: Hasegawa, So;Kitagawa, Ryoma;Hasegawa, Mikio
• 通讯作者: Hasegawa, Mikio