[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126469-en":3,"doc-seo-126469-105":31,"detail-sidebar-cat-0-en-105":93},{"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},126469,962084925290,"Ophelia","https://ap-avatar.wpscdn.com/davatar_085a072bc5b1113ac321206ff7593b45",8,"Research & Report","Scalable Machine-Learning-Based Classification of Radio Frequency Building Loss - Master’s Thesis 2025","Accurately modeling outdoor-to-indoor (O2I) and indoor-to-indoor (I2I) signal loss is crucial for enhancing indoor wireless network performance in dense urban environments. Traditional on-site measurement is costly and hard to scale, while real datasets often show noise and class imbalance that hinder reliable prediction. This thesis presents a scalable machine-learning framework combining passive crowdsourced UE measurements from 3GPP-compliant networks with building attributes from public datasets, then evaluates multiple supervised and semi-supervised ensemble models and post-analyzes robustness across frequency bands and building types.","Master’s Programme in Master’s Programme in ICT Innovation  \nScalable Machine-Learning-Based Classification of Radio Frequency Building Loss  \nJiayi Tan  \nMaster’s Thesis 2025  \n© 2025  \nThis work is licensed under a Creative Commons“Attribution-NonCommercial-ShareAlike 4 .0 International” license.  \n| Author Jiayi Tan |\n| --- |\n| Title Scalable Machine-Learning-Based Classification of Radio Frequency Building Loss |\n| Degree programme Master’s Programme in ICT Innovation |\n| Major Autonomous Systems |\n| Supervisor Prof. Quan Zhou |\n| Advisors James Gross, Neelabhro Roy |\n| Collaborative partner Ericsson AB |\n| Date 15 November 2025 Number of pages 78+18 Language English |\n| Abstract\u003Cbr>Accurately modeling outdoor-to-indoor (O2I) and indoor-to-indoor (I2I) signal loss is crucial for enhancing indoor wireless network performance in dense urban environments. Traditional on-site measurement methods are costly, time-consuming, and difficult to scale. Real-world datasets often suffer from noise and imbalance, which complicates reliable signal loss prediction. Building on prior work such as ITU-R P.1238 and the COST 231 Multi-Wall (COST231) model, this study proposed a scalable machine learning framework for classifying radio frequency (RF) building loss. The framework integrated passively collected, crowdsourced user equipment (UE) measurements from 3GPP-compliant networks. These were combined with building attributes from public datasets. We evaluated Random Forest, XGBoost, LightGBM, and an integrated voting classifier. Analyses included both supervised and semi-supervised learning approaches. The output loss classifications were further post-analyzed alongside the input RF and building features to verify robustness and validity. This assessment covered diverse frequency bands and building types, based on empirical observations. Results showed that all models performed well in classifying building loss. Notably, the semi-supervised XGBoost model achieved the best performance for O2I classification. Semi-supervised LightGBM excelled in I2I classification. This approach demonstrates a scalable, data-driven alternative to traditional methods and offers actionable insights for network planning and indoor coverage optimization. |\n| Keywords Building Loss, Outdoor-to-Indoor (O2I), Indoor-to-Indoor (I2I), 3GPP, Machine Learning, OpenStreetMap (OSM) |\n\nPreface  \nI would like to thank my supervisors, Rohit Chandra, and Neelabhro Roy, for their support, guidance, and helpful feedback throughout my internship. I also want to thank my examiner, James Gross, for reviewing my work and providing valuable comments.  \nIn addition, special thanks to Jason Chen, and Billy Hogan, for their advice and help during the project. I am also grateful to everyone at Ericsson for making me feel welcome and for their support during my time there.  \nFurthermore, I want to thank my family, and my friends. Without their support and encouragement, I would not have made it this far.  \nLastly, I would also like to thank everyone who has generously helped me throughout my years of study. Although I cannot list all your names here, please know that I sincerely appreciate your support from the bottom of my heart.  \nOtaniemi, 30 September 2025  \nJiayi Tan  \nContents  \nAbstract 3  \nPreface 4  \nContents 5  \nAbbreviations 9  \n1 Introduction 11  \n1.1 Background .............................. 11  \n1.2 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12  \n1.2.1 Challenges in Collecting Network Measurement Data (O2I & I2I) .............................. 13  \n1.2.2 Limitations in Modeling Building Characteristics ...... 13  \n1.2.3 Scalability and Accuracy Issues in Network Planning Tools . 14  \n1.3 Motivation & Purpose ......................... 14  \n1.4 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . 15  \n1.5 Research Methodology . . . . . . . . . . . . . . . . . . . . . . . . 16  \n1.6 Delimitations . . . . . . . . . . . . . . . . ","cbCaifgtgF4WiUWw","https://ap.wps.com/l/cbCaifgtgF4WiUWw","pdf",15374778,7,1,95,"English","en",105,"# Abstract\n# Preface\n# Contents\n# Abbreviations\n# 1 Introduction\n## 1.1 Background\n## 1.2 Problem\n## 1.3 Motivation & Purpose\n## 1.4 Research Questions\n## 1.5 Research Methodology\n## 1.6 Delimitations\n## 1.7 Structure of the thesis\n# 2 Background\n## 2.1 RF Signal Propagation\n## 2.2 Propagation Loss Models\n## 2.3 Ensemble ML Models\n## 2.4 Energy Efficiency of Buildings & RF Signal Loss\n## 2.5 Summary\n# 3 Methods\n## 3.1 Overview\n## 3.2 Data\n## 3.3 Data Preprocessing & Feature Engineering","[{\"question\":\"Why is modeling O2I and I2I signal loss important in dense urban environments?\",\"answer\":\"It directly affects indoor wireless network performance and coverage. Accurate modeling supports better network planning and indoor optimization.\"},{\"question\":\"What data sources does the proposed framework use?\",\"answer\":\"It combines passively collected, crowdsourced UE measurements from 3GPP-compliant networks with building attributes from public datasets.\"},{\"question\":\"Which models performed best for O2I and I2I classification?\",\"answer\":\"Semi-supervised XGBoost achieved the best O2I performance, while semi-supervised LightGBM excelled for I2I classification.\"}]","Scalable Machine-Learning-Based Classification of Radio Frequency Building Loss - Master’s Thesis 2025 | PDF",1785905227,239,{"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":88,"head_meta":90,"extra_data":92,"updated_unix":29},"scalable-machine-learning-based-classification-of-radio-frequency-building-loss-masters-thesis-2025","",{"@graph":37,"@context":87},[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/research-report/",3,{"item":53,"name":13,"@type":44,"position":54},"https://docshare.wps.com/document/scalable-machine-learning-based-classification-of-radio-frequency-building-loss-masters-thesis-2025/126469/",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],{"name":74,"@type":75,"acceptedAnswer":76},"Why is modeling O2I and I2I signal loss important in dense urban environments?","Question",{"text":77,"@type":78},"It directly affects indoor wireless network performance and coverage. Accurate modeling supports better network planning and indoor optimization.","Answer",{"name":80,"@type":75,"acceptedAnswer":81},"What data sources does the proposed framework use?",{"text":82,"@type":78},"It combines passively collected, crowdsourced UE measurements from 3GPP-compliant networks with building attributes from public datasets.",{"name":84,"@type":75,"acceptedAnswer":85},"Which models performed best for O2I and I2I classification?",{"text":86,"@type":78},"Semi-supervised XGBoost achieved the best O2I performance, while semi-supervised LightGBM excelled for I2I classification.","https://schema.org",{"og:url":53,"og:type":89,"og:title":13,"og:site_name":60,"og:description":14},"article",{"robots":91,"canonical":53},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":94},[95,99,103,107,112,117,121,124,129,132,136],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":96,"show_sort_weight":97,"slug":98},"Story & Novel",90,"story-novel",{"id":48,"doc_module":4,"doc_module_name":47,"category_name":100,"show_sort_weight":101,"slug":102},"Literature",80,"literature",{"id":54,"doc_module":4,"doc_module_name":47,"category_name":104,"show_sort_weight":105,"slug":106},"Exam",70,"exam",{"id":108,"doc_module":4,"doc_module_name":47,"category_name":109,"show_sort_weight":110,"slug":111},5,"Comic",60,"comic",{"id":113,"doc_module":4,"doc_module_name":47,"category_name":114,"show_sort_weight":115,"slug":116},6,"Technology",50,"technology",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":118,"show_sort_weight":119,"slug":120},"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":47,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":47,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":47,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":47,"category_name":138,"show_sort_weight":108,"slug":139},19,"General","general"]