集成信道估计的智能反射面先进波束形成与反射控制

IF 1.1 4区 计算机科学 Q4 ENGINEERING, ELECTRICAL & ELECTRONIC
Sakhshra Monga, Anmol Rattan Singh, Nitin Saluja, Chander Prabha, Shivani Malhotra, Asif Karim, Md. Mehedi Hassan
{"title":"集成信道估计的智能反射面先进波束形成与反射控制","authors":"Sakhshra Monga,&nbsp;Anmol Rattan Singh,&nbsp;Nitin Saluja,&nbsp;Chander Prabha,&nbsp;Shivani Malhotra,&nbsp;Asif Karim,&nbsp;Md. Mehedi Hassan","doi":"10.1049/mia2.12538","DOIUrl":null,"url":null,"abstract":"<p>Intelligent Reflecting Surfaces (IRS) enhance wireless communication by optimising signal reflection from the base station (BS) towards users. The passive nature of IRS components makes tuning phase shifters difficult and direct channel measurement problematic. This study presents a machine learning framework that directly maximises the beamformers at the BS and the reflective coefficients at the IRS, bypassing conventional methods that estimate channels before optimising system parameters. This is achieved by mapping incoming pilot signals and data, including user positions, with a deep neural network (DNN), guiding an optimal setup. User interactions are captured using a permutation-invariant graph neural network (GNN) architecture. Simulation results show that implicit channel estimation method requires fewer pilots than standard approaches, effectively learns to optimise sum rate or minimum-rate targets, and generalises well. Specifically, the sum rate for GDNNet (GNN + DNN) improves by <span></span><math>\n <semantics>\n <mrow>\n <mn>12.57</mn>\n <mi>%</mi>\n </mrow>\n <annotation> $12.57\\%$</annotation>\n </semantics></math> over linear minimum mean square error (LMMSE) and by <span></span><math>\n <semantics>\n <mrow>\n <mn>12.42</mn>\n <mi>%</mi>\n </mrow>\n <annotation> $12.42\\%$</annotation>\n </semantics></math> over perfect CSI concerning the number of users, and by <span></span><math>\n <semantics>\n <mrow>\n <mn>28.57</mn>\n <mi>%</mi>\n </mrow>\n <annotation> $28.57\\%$</annotation>\n </semantics></math> over LMMSE and by <span></span><math>\n <semantics>\n <mrow>\n <mn>14.28</mn>\n <mi>%</mi>\n </mrow>\n <annotation> $14.28\\%$</annotation>\n </semantics></math> over perfect CSI concerning pilot length. Offering a feasible solution with reduced computing complexity for real-world applications, the proposed GNN + DNN method outperforms conventional model-based techniques such as LMMSE and approaches the performance of perfect CSI, demonstrating its high effectiveness in various scenarios.</p>","PeriodicalId":13374,"journal":{"name":"Iet Microwaves Antennas & Propagation","volume":"18 12","pages":"917-931"},"PeriodicalIF":1.1000,"publicationDate":"2024-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/mia2.12538","citationCount":"0","resultStr":"{\"title\":\"Advanced beamforming and reflection control in intelligent reflecting surfaces with integrated channel estimation\",\"authors\":\"Sakhshra Monga,&nbsp;Anmol Rattan Singh,&nbsp;Nitin Saluja,&nbsp;Chander Prabha,&nbsp;Shivani Malhotra,&nbsp;Asif Karim,&nbsp;Md. Mehedi Hassan\",\"doi\":\"10.1049/mia2.12538\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>Intelligent Reflecting Surfaces (IRS) enhance wireless communication by optimising signal reflection from the base station (BS) towards users. The passive nature of IRS components makes tuning phase shifters difficult and direct channel measurement problematic. This study presents a machine learning framework that directly maximises the beamformers at the BS and the reflective coefficients at the IRS, bypassing conventional methods that estimate channels before optimising system parameters. This is achieved by mapping incoming pilot signals and data, including user positions, with a deep neural network (DNN), guiding an optimal setup. User interactions are captured using a permutation-invariant graph neural network (GNN) architecture. Simulation results show that implicit channel estimation method requires fewer pilots than standard approaches, effectively learns to optimise sum rate or minimum-rate targets, and generalises well. Specifically, the sum rate for GDNNet (GNN + DNN) improves by <span></span><math>\\n <semantics>\\n <mrow>\\n <mn>12.57</mn>\\n <mi>%</mi>\\n </mrow>\\n <annotation> $12.57\\\\%$</annotation>\\n </semantics></math> over linear minimum mean square error (LMMSE) and by <span></span><math>\\n <semantics>\\n <mrow>\\n <mn>12.42</mn>\\n <mi>%</mi>\\n </mrow>\\n <annotation> $12.42\\\\%$</annotation>\\n </semantics></math> over perfect CSI concerning the number of users, and by <span></span><math>\\n <semantics>\\n <mrow>\\n <mn>28.57</mn>\\n <mi>%</mi>\\n </mrow>\\n <annotation> $28.57\\\\%$</annotation>\\n </semantics></math> over LMMSE and by <span></span><math>\\n <semantics>\\n <mrow>\\n <mn>14.28</mn>\\n <mi>%</mi>\\n </mrow>\\n <annotation> $14.28\\\\%$</annotation>\\n </semantics></math> over perfect CSI concerning pilot length. Offering a feasible solution with reduced computing complexity for real-world applications, the proposed GNN + DNN method outperforms conventional model-based techniques such as LMMSE and approaches the performance of perfect CSI, demonstrating its high effectiveness in various scenarios.</p>\",\"PeriodicalId\":13374,\"journal\":{\"name\":\"Iet Microwaves Antennas & Propagation\",\"volume\":\"18 12\",\"pages\":\"917-931\"},\"PeriodicalIF\":1.1000,\"publicationDate\":\"2024-12-07\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://onlinelibrary.wiley.com/doi/epdf/10.1049/mia2.12538\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Iet Microwaves Antennas & Propagation\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1049/mia2.12538\",\"RegionNum\":4,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Iet Microwaves Antennas & Propagation","FirstCategoryId":"94","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1049/mia2.12538","RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0

摘要

智能反射面(IRS)通过优化从基站(BS)到用户的信号反射来增强无线通信。IRS元件的被动特性使得移相器的调谐变得困难,直接通道测量也存在问题。本研究提出了一种机器学习框架,可以直接最大化BS处的波束形成器和IRS处的反射系数,绕过了在优化系统参数之前估计通道的传统方法。这是通过映射输入的导频信号和数据来实现的,包括用户位置,并使用深度神经网络(DNN)来指导最佳设置。使用排列不变图神经网络(GNN)架构捕获用户交互。仿真结果表明,隐式信道估计方法比标准方法需要更少的导频,能够有效地学习和速率或最小速率目标的优化,具有良好的泛化性。具体来说,GDNNet (GNN + DNN)的总和率比线性最小均方误差(LMMSE)提高了12.57%,比完美CSI在用户数量上提高了12.42%。比LMMSE高28.57%,比完美CSI高14.28%。提出的GNN + DNN方法为实际应用提供了一种可行的解决方案,降低了计算复杂性,优于传统的基于模型的技术,如LMMSE,并接近完美的CSI性能,在各种场景中显示出其高效率。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Advanced beamforming and reflection control in intelligent reflecting surfaces with integrated channel estimation

Advanced beamforming and reflection control in intelligent reflecting surfaces with integrated channel estimation

Intelligent Reflecting Surfaces (IRS) enhance wireless communication by optimising signal reflection from the base station (BS) towards users. The passive nature of IRS components makes tuning phase shifters difficult and direct channel measurement problematic. This study presents a machine learning framework that directly maximises the beamformers at the BS and the reflective coefficients at the IRS, bypassing conventional methods that estimate channels before optimising system parameters. This is achieved by mapping incoming pilot signals and data, including user positions, with a deep neural network (DNN), guiding an optimal setup. User interactions are captured using a permutation-invariant graph neural network (GNN) architecture. Simulation results show that implicit channel estimation method requires fewer pilots than standard approaches, effectively learns to optimise sum rate or minimum-rate targets, and generalises well. Specifically, the sum rate for GDNNet (GNN + DNN) improves by 12.57 % $12.57\%$ over linear minimum mean square error (LMMSE) and by 12.42 % $12.42\%$ over perfect CSI concerning the number of users, and by 28.57 % $28.57\%$ over LMMSE and by 14.28 % $14.28\%$ over perfect CSI concerning pilot length. Offering a feasible solution with reduced computing complexity for real-world applications, the proposed GNN + DNN method outperforms conventional model-based techniques such as LMMSE and approaches the performance of perfect CSI, demonstrating its high effectiveness in various scenarios.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Iet Microwaves Antennas & Propagation
Iet Microwaves Antennas & Propagation 工程技术-电信学
CiteScore
4.30
自引率
5.90%
发文量
109
审稿时长
7 months
期刊介绍: Topics include, but are not limited to: Microwave circuits including RF, microwave and millimetre-wave amplifiers, oscillators, switches, mixers and other components implemented in monolithic, hybrid, multi-chip module and other technologies. Papers on passive components may describe transmission-line and waveguide components, including filters, multiplexers, resonators, ferrite and garnet devices. For applications, papers can describe microwave sub-systems for use in communications, radar, aerospace, instrumentation, industrial and medical applications. Microwave linear and non-linear measurement techniques. Antenna topics including designed and prototyped antennas for operation at all frequencies; multiband antennas, antenna measurement techniques and systems, antenna analysis and design, aperture antenna arrays, adaptive antennas, printed and wire antennas, microstrip, reconfigurable, conformal and integrated antennas. Computational electromagnetics and synthesis of antenna structures including phased arrays and antenna design algorithms. Radiowave propagation at all frequencies and environments. Current Special Issue. Call for papers: Metrology for 5G Technologies - https://digital-library.theiet.org/files/IET_MAP_CFP_M5GT_SI2.pdf
×
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