Underlay Cognitive Radio Resource Management with Hybrid Meta-Loss Learning

IF 1.5 4区 工程技术 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC
Nikita Mishra, Sumit Srivastava, Shivendra Nath Sharan
{"title":"Underlay Cognitive Radio Resource Management with Hybrid Meta-Loss Learning","authors":"Nikita Mishra, Sumit Srivastava, Shivendra Nath Sharan","doi":"10.1007/s40998-024-00733-x","DOIUrl":null,"url":null,"abstract":"<p>Cognitive Radio (CR) is an adaptable communication device driven by a Cognitive Engine (CE). A suitable machine-learning strategy can increase the learning potential of CE. This work proposes an underlay Resource Allocation (RA) framework for maximizing cognitive user Quality of Experience (QoE) while simultaneously protecting the primary network. A non-convex optimization problem is resolved using a combination of two algorithms: Meta Actor-Critic (MAC) loss and Deep Deterministic Policy Gradient (DDPG). The incorporation of the MAC neural network into RA improves the actor’s performance, which in turn boosts the entire system’s learning speed. The proposed framework is validated through a comparative analysis with four contemporary RA models: Q learning, hybrid, dueling DQN, and MAML. The simulation findings exemplify that the proposed method achieves convergence rapidly and improves reward value in a shorter amount of time than the other four extant RA models. Further, the impact of increasing system size on the distortion versus scalability trade-off is also highlighted.</p>","PeriodicalId":49064,"journal":{"name":"Iranian Journal of Science and Technology-Transactions of Electrical Engineering","volume":null,"pages":null},"PeriodicalIF":1.5000,"publicationDate":"2024-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Iranian Journal of Science and Technology-Transactions of Electrical Engineering","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1007/s40998-024-00733-x","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0

Abstract

Cognitive Radio (CR) is an adaptable communication device driven by a Cognitive Engine (CE). A suitable machine-learning strategy can increase the learning potential of CE. This work proposes an underlay Resource Allocation (RA) framework for maximizing cognitive user Quality of Experience (QoE) while simultaneously protecting the primary network. A non-convex optimization problem is resolved using a combination of two algorithms: Meta Actor-Critic (MAC) loss and Deep Deterministic Policy Gradient (DDPG). The incorporation of the MAC neural network into RA improves the actor’s performance, which in turn boosts the entire system’s learning speed. The proposed framework is validated through a comparative analysis with four contemporary RA models: Q learning, hybrid, dueling DQN, and MAML. The simulation findings exemplify that the proposed method achieves convergence rapidly and improves reward value in a shorter amount of time than the other four extant RA models. Further, the impact of increasing system size on the distortion versus scalability trade-off is also highlighted.

Abstract Image

利用混合元损失学习的下层认知无线电资源管理
认知无线电(CR)是一种由认知引擎(CE)驱动的适应性通信设备。合适的机器学习策略可以提高认知引擎的学习潜力。本研究提出了一种底层资源分配(RA)框架,用于最大限度地提高认知用户的体验质量(QoE),同时保护主网络。结合两种算法解决了一个非凸优化问题:元行为批判 (MAC) 损失和深度确定性策略梯度 (DDPG)。在 RA 中加入 MAC 神经网络可提高行为体的性能,进而提高整个系统的学习速度。通过与四种当代 RA 模型的对比分析,验证了所提出的框架:Q 学习、混合、决斗 DQN 和 MAML。仿真结果表明,与其他四种现存 RA 模型相比,所提出的方法能快速收敛并在更短的时间内提高奖励值。此外,还强调了系统规模增大对失真与可扩展性权衡的影响。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
CiteScore
5.50
自引率
4.20%
发文量
93
审稿时长
>12 weeks
期刊介绍: Transactions of Electrical Engineering is to foster the growth of scientific research in all branches of electrical engineering and its related grounds and to provide a medium by means of which the fruits of these researches may be brought to the attentionof the world’s scientific communities. The journal has the focus on the frontier topics in the theoretical, mathematical, numerical, experimental and scientific developments in electrical engineering as well as applications of established techniques to new domains in various electical engineering disciplines such as: Bio electric, Bio mechanics, Bio instrument, Microwaves, Wave Propagation, Communication Theory, Channel Estimation, radar & sonar system, Signal Processing, image processing, Artificial Neural Networks, Data Mining and Machine Learning, Fuzzy Logic and Systems, Fuzzy Control, Optimal & Robust ControlNavigation & Estimation Theory, Power Electronics & Drives, Power Generation & Management The editors will welcome papers from all professors and researchers from universities, research centers, organizations, companies and industries from all over the world in the hope that this will advance the scientific standards of the journal and provide a channel of communication between Iranian Scholars and their colleague in other parts of the world.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信