[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126342-en":3,"doc-seo-126342-105":31,"detail-sidebar-cat-0-en-105":97},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},126342,962085570644,"Evangeline","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",6,"Technology","High-speed Machine Learning-enhanced Receiver for Millimeter-Wave Systems","Machine Learning enables the design of wireless physical-layer components, especially for millimeter-wave systems where hardware imperfections and rapidly varying channels challenge conventional processing. This work proposes a full ML-enhanced mm-wave receiver that jointly performs channel estimation, equalization, phase correction, and demapping with convolutional neural networks for frequency-selective channels. It introduces a guard-interval-based channel refresh to address short-timescale tracking needs. Simulations show up to 7 dB gains, and a 60 GHz FPGA testbed yields throughput up to 6× in mobile scenarios.","High-speed Machine Learning-enhanced Receiver for Millimeter-Wave Systems  \nDolores Garcia†⋆, Rafael Ruiz†⋆, Jesus O. Lacruz†, Joerg Widmer†  \n† IMDEA Networks Institute, Madrid, Spain  \n⋆ Universidad Carlos III de Madrid, Madrid, Spain  \nE-mails:† {[firstname.lastname](firstname.lastname}@imdea.org)[}](firstname.lastname}@imdea.org)[@imdea.org](firstname.lastname}@imdea.org)  \nAbstract—Machine Learning (ML) is a promising tool to design wireless physical layer (PHY) components. It is particularly interesting for millimeter-wave (mm-wave) frequencies and above, due to the more challenging hardware design and channel environment at these frequencies. Rather than building individual ML-components, in this paper, we design an entire ML-enhanced mm-wave receiver for frequency selective channels. Our ML-receiver jointly optimizes the channel estimation, equalization, phase correction and demapper using Convolutional Neural Networks. We also show that for mm-wave systems, the channel varies significantly even over short timescales, requiring frequent channel measurements, and this situation is exacerbated in mobile scenarios. To tackle this, we propose a new MLchannel estimation approach that refreshes the channel state information using the guard intervals (not intended for channel measurements) that are available for every block of symbols in communication packets. To the best of our knowledge, our MLreceiver is the first work to outperform conventional receiversin general scenarios, with simulation results showing up to 7 dB gains. We also provide an experimental validation of the ML-enhanced receiver with a 60 GHz FPGA-based testbed with phased antenna arrays, which shows a throughput increase by a factor of up to 6 over baseline schemes in mobile scenarios.  \nIndex Terms—Machine Learning, Millimeter Wave, Channel Estimation, Physical Layer, Equalization  \nI. INTRODUCTION  \nMachine Learning (ML) technology has shown great advances over the last decade, mainly due to the abundant and affordable high-performance computation engines that allow fast training of large neural networks. In wireless systems, ML has been extensively used at high network protocol layers, for example for traffic prediction, blockage prediction or anomaly detection [1–3], where ML techniques often outperform classical algorithms. Recently, several works showcased that ML is also suitable to design components of the wireless physical layer (PHY) [4–12] . Their main motivation is that conventional PHY layers are composed of highly optimized but separate signal processing blocks such as modulator/demodulator, equalizer, demapper, etc. This block-wise design facilitates implementation and modularity, but does not always result in optimum performance. Furthermore, these blocks are often designed under the assumption of linear processes without considering potential non-linearities and residual errors due to the interaction among blocks. Also channel tracking over short timescales is usually not a main concern in these designs. Such issues become more relevant at higher frequencies, in  \nparticular the millimeter-wave (mm-wave) band and at THz frequencies, where the high frequency induces channel fluctuations even in static scenarios, device complexity increases, and miniaturization makes hardware imperfections more complex and harder to model.  \nDesigning high-performance RF components is more complex at mm-wave frequencies, and ML-based systems can inherently learn how to cope with imperfections. Achieving high gains with small mm-wave components is difficult, and they are more susceptible to noise and deviations. Giga-sampling rate AD/DA converters for wide-band channels often have limited resolution to avoid excessive complexity and power consumption, which introduces non-negligible quantization effects [13] . The directional antenna arrays used in mm-wave systems to compensate for the high path loss require complex calibration methods to ensure matching betw","cbCaij2sWit8ciHC","https://ap.wps.com/l/cbCaij2sWit8ciHC","pdf",5332478,9,1,10,"English","en",105,"# Introduction\n## Machine Learning for wireless physical-layer design\n## Limitations of block-wise conventional PHY processing\n## Motivation for millimeter-wave and THz systems\n## Prior ML receiver architectures and gaps\n## Proposed single-carrier ML-enhanced receiver (ML4RX)","[{\"question\":\"为什么将机器学习用于毫米波物理层接收机更具意义？\",\"answer\":\"毫米波频段硬件更难设计且信道会在短时间尺度内显著波动，传统分块式信号处理在存在非线性与残余误差时难以达到最优。\"},{\"question\":\"论文中的ML增强接收机具体做哪些任务？\",\"answer\":\"它用卷积神经网络联合优化信道估计、均衡、相位校正以及解调映射（demapper）。\"},{\"question\":\"如何缓解毫米波场景下需要频繁信道测量的问题？\",\"answer\":\"提出一种新的信道估计方法，利用每个符号块中的保护间隔（原本不用于信道测量）刷新信道状态信息。\"},{\"question\":\"实验结果如何验证该方法的有效性？\",\"answer\":\"在60 GHz、基于FPGA的相控阵测试平台上，移动场景相较基线方案吞吐量最高可提升6倍；仿真也显示最高约7 dB增益。\"}]","High-speed Machine Learning-enhanced Receiver for Millimeter-Wave Systems | PDF",1785904566,25,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":92,"head_meta":94,"extra_data":96,"updated_unix":29},"high-speed-machine-learning-enhanced-receiver-for-millimeter-wave-systems","",{"@graph":37,"@context":91},[38,55,70],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,49,52],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":48},"https://docshare.wps.com/document/","Document",2,{"item":50,"name":12,"@type":44,"position":51},"https://docshare.wps.com/document/technology/",3,{"item":53,"name":13,"@type":44,"position":54},"https://docshare.wps.com/document/high-speed-machine-learning-enhanced-receiver-for-millimeter-wave-systems/126342/",4,{"url":53,"name":13,"@type":56,"author":57,"headline":13,"publisher":59,"fileFormat":62,"inLanguage":24,"description":14,"dateModified":63,"datePublished":64,"encodingFormat":62,"isAccessibleForFree":65,"interactionStatistic":66},"DigitalDocument",{"name":9,"@type":58},"Person",{"url":42,"name":60,"@type":61},"DocShare","Organization","application/pdf","2026-08-22","2026-08-05",true,{"@type":67,"interactionType":68,"userInteractionCount":20},"InteractionCounter",{"@type":69},"ViewAction",{"@type":71,"mainEntity":72},"FAQPage",[73,79,83,87],{"name":74,"@type":75,"acceptedAnswer":76},"为什么将机器学习用于毫米波物理层接收机更具意义？","Question",{"text":77,"@type":78},"毫米波频段硬件更难设计且信道会在短时间尺度内显著波动，传统分块式信号处理在存在非线性与残余误差时难以达到最优。","Answer",{"name":80,"@type":75,"acceptedAnswer":81},"论文中的ML增强接收机具体做哪些任务？",{"text":82,"@type":78},"它用卷积神经网络联合优化信道估计、均衡、相位校正以及解调映射（demapper）。",{"name":84,"@type":75,"acceptedAnswer":85},"如何缓解毫米波场景下需要频繁信道测量的问题？",{"text":86,"@type":78},"提出一种新的信道估计方法，利用每个符号块中的保护间隔（原本不用于信道测量）刷新信道状态信息。",{"name":88,"@type":75,"acceptedAnswer":89},"实验结果如何验证该方法的有效性？",{"text":90,"@type":78},"在60 GHz、基于FPGA的相控阵测试平台上，移动场景相较基线方案吞吐量最高可提升6倍；仿真也显示最高约7 dB增益。","https://schema.org",{"og:url":53,"og:type":93,"og:title":13,"og:site_name":60,"og:description":14},"article",{"robots":95,"canonical":53},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":98},[99,103,107,111,116,119,124,129,133,136,139],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":100,"show_sort_weight":101,"slug":102},"Story & Novel",90,"story-novel",{"id":48,"doc_module":4,"doc_module_name":47,"category_name":104,"show_sort_weight":105,"slug":106},"Literature",80,"literature",{"id":54,"doc_module":4,"doc_module_name":47,"category_name":108,"show_sort_weight":109,"slug":110},"Exam",70,"exam",{"id":112,"doc_module":4,"doc_module_name":47,"category_name":113,"show_sort_weight":114,"slug":115},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":117,"slug":118},50,"technology",{"id":120,"doc_module":4,"doc_module_name":47,"category_name":121,"show_sort_weight":122,"slug":123},7,"Healthcare",40,"healthcare",{"id":125,"doc_module":4,"doc_module_name":47,"category_name":126,"show_sort_weight":127,"slug":128},8,"Research & Report",30,"research-report",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":130,"show_sort_weight":131,"slug":132},"Religion & Spirituality",20,"religion-spirituality",{"id":131,"doc_module":4,"doc_module_name":47,"category_name":134,"show_sort_weight":131,"slug":135},"World Cup","world-cup",{"id":22,"doc_module":4,"doc_module_name":47,"category_name":137,"show_sort_weight":22,"slug":138},"Lifestyle","lifestyle",{"id":140,"doc_module":4,"doc_module_name":47,"category_name":141,"show_sort_weight":112,"slug":142},19,"General","general"]