{"id":{"repo_id":"msu","oai_identifier":"oai:d.lib.msu.edu:etd_53026"},"canonical_url":"https://search.ndltd.org/etd/msu/oai:d.lib.msu.edu:etd_53026","repository":{"repo_id":"msu","name":"Michigan State University","base_url":"https://d.lib.msu.edu/oai"},"display":{"title":"TOWARD COOPERATIVE EDGE INTELLIGENCE : INTEGRATING LEARNING AND SENSING IN INTERNET-OF-THINGS SYSTEMS","abstract":"Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2026","abstract_html":"Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2026","abstract_has_math":false,"creators":["Gan, Maolin"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Cao, Zhichao","Zeng, Huacheng","Li, Tianxing","Xie, Yuying"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026","date_published":"2026","updated_at":"2026-11-18T18:59:37Z","subjects":["Computer science"],"languages":["eng"],"rights":["In Copyright"],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:53026","local:Gan_grad.msu_0128D_22030"],"render_values":[{"text":"etd:53026","href":null,"code":true},{"text":"local:Gan_grad.msu_0128D_22030","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://doi.org/doi:10.25335/kb01-at75","outbound_label":"DOI","outbound_source":"dc:identifier","record_url":"https://doi.org/doi:10.25335/kb01-at75","record_label":"Repository record","record_kind":"doi","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://d.lib.msu.edu/oai?verb=GetRecord&metadataPrefix=oai_dc&identifier=oai%3Ad.lib.msu.edu%3Aetd_53026","prefix":"oai_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Cao, Zhichao","Zeng, Huacheng","Li, Tianxing","Xie, Yuying"]},{"key":"dc:creator","label":"Author","values":["Gan, Maolin"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2026"]},{"key":"dc:relation","label":"Dc Relation","values":["Electronic Theses & Dissertations"]},{"key":"dc:type","label":"Dc Type","values":["text","Electronic dissertations"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Computer science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["In Copyright"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:53026","local:Gan_grad.msu_0128D_22030","https://doi.org/doi:10.25335/kb01-at75"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2026","The proliferation of Internet-of-Things (IoT) systems is reshaping a wide range of applications by enabling pervasive sensing, real-time data collection, and intelligent decision-making. Traditionally, IoT systems have relied on cloud-centric architectures, which suffer from latency, bandwidth limitations, privacy risks, and scalability issues. These limitations have motivated the shift toward edge intelligence, where sensing, learning, and computation are performed closer to data sources. Realizing such efficient and cooperative edge intelligence in practice, however, remains challenging because IoT systems are inherently distributed, resource-constrained, heterogeneous, and often deployed under noisy sensing conditions and unstable network connectivity, necessitating novel solutions for IoT-edge integration.This dissertation advances cooperative edge intelligence tailored to IoT systems through two complementary dimensions, which are distributed edge learning and intelligent edge sensing, with the goal of enabling more efficient, accurate, and adaptive IoT systems. Distributed edge learning focuses on decentralized model training across edge devices without sharing raw data, where Federated Learning serves as a representative paradigm. It addresses key challenges in real-world IoT deployments, including communication efficiency, statistical and system heterogeneity, and model divergence caused by non–Independent and Identically Distributed (non-IID) data distributions. Intelligent edge sensing, on the other hand, enhances the perceptual capabilities of edge devices by leveraging signal processing and machine learning techniques to extract meaningful, task-relevant information from raw sensor data. This enables robust and fine-grained perception under dynamic and noisy environments. Integrating these two dimensions forms the foundation of cooperative edge intelligence, where efficient collaboration in learning and high-quality perception in sensing support end-to-end IoT intelligence. On the distributed learning side, Chapter 2 presents GeoFL, which is a hierarchical federated learning framework for geo-distributed IoT systems, addressing the challenges of bandwidth-constrained wide-area networks, aggregator-level heterogeneity, and stale model updates through importance-aware communication, adaptive aggregation, and graph-based modeling. On the intelligent sensing side, Chapter 3 introduces Adonis, a neural-enhanced mmWave sensing system that enables fine-grained and robust leaf wetness perception in agricultural IoT, leveraging multi-perspective signal representations and lightweight learning models to extract task-relevant features under noisy environmental conditions. Chapter 4 proposes 3D-FAD, which integrates distributed learning and intelligent sensing in a unified multimodal federated framework for autonomous driving, where modality informativeness is explicitly modeled to guide feature fusion, client contribution, and masked modality autoencoder is introduced to compensate for deficient LiDAR field of view using correlated camera information. These systems demonstrate how distributed learning and intelligent sensing can be co-designed to reinforce each other, enabling IoT systems to collaboratively optimize both data acquisition and model learning under real-world constraints. Taken together, this dissertation establishes a unified system perspective on cooperative edge intelligence through learning and sensing. By bridging communication-efficient federated learning, neural-enhanced edge sensing, and multimodal federated perception, this dissertation provides new system and algorithmic foundations for building scalable, adaptive, and intelligent IoT systems.","Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).","Includes bibliographical references."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf","1 online resource (138 p.) : ill."]},{"key":"dc:title","label":"Title","values":["TOWARD COOPERATIVE EDGE INTELLIGENCE : INTEGRATING LEARNING AND SENSING IN INTERNET-OF-THINGS SYSTEMS"]}]}],"canonical_facts":{"dc:contributor":["Cao, Zhichao","Zeng, Huacheng","Li, Tianxing","Xie, Yuying"],"dc:creator":["Gan, Maolin"],"dc:date":["2026"],"dc:description":["Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2026","The proliferation of Internet-of-Things (IoT) systems is reshaping a wide range of applications by enabling pervasive sensing, real-time data collection, and intelligent decision-making. Traditionally, IoT systems have relied on cloud-centric architectures, which suffer from latency, bandwidth limitations, privacy risks, and scalability issues. These limitations have motivated the shift toward edge intelligence, where sensing, learning, and computation are performed closer to data sources. Realizing such efficient and cooperative edge intelligence in practice, however, remains challenging because IoT systems are inherently distributed, resource-constrained, heterogeneous, and often deployed under noisy sensing conditions and unstable network connectivity, necessitating novel solutions for IoT-edge integration.This dissertation advances cooperative edge intelligence tailored to IoT systems through two complementary dimensions, which are distributed edge learning and intelligent edge sensing, with the goal of enabling more efficient, accurate, and adaptive IoT systems. Distributed edge learning focuses on decentralized model training across edge devices without sharing raw data, where Federated Learning serves as a representative paradigm. It addresses key challenges in real-world IoT deployments, including communication efficiency, statistical and system heterogeneity, and model divergence caused by non–Independent and Identically Distributed (non-IID) data distributions. Intelligent edge sensing, on the other hand, enhances the perceptual capabilities of edge devices by leveraging signal processing and machine learning techniques to extract meaningful, task-relevant information from raw sensor data. This enables robust and fine-grained perception under dynamic and noisy environments. Integrating these two dimensions forms the foundation of cooperative edge intelligence, where efficient collaboration in learning and high-quality perception in sensing support end-to-end IoT intelligence. On the distributed learning side, Chapter 2 presents GeoFL, which is a hierarchical federated learning framework for geo-distributed IoT systems, addressing the challenges of bandwidth-constrained wide-area networks, aggregator-level heterogeneity, and stale model updates through importance-aware communication, adaptive aggregation, and graph-based modeling. On the intelligent sensing side, Chapter 3 introduces Adonis, a neural-enhanced mmWave sensing system that enables fine-grained and robust leaf wetness perception in agricultural IoT, leveraging multi-perspective signal representations and lightweight learning models to extract task-relevant features under noisy environmental conditions. Chapter 4 proposes 3D-FAD, which integrates distributed learning and intelligent sensing in a unified multimodal federated framework for autonomous driving, where modality informativeness is explicitly modeled to guide feature fusion, client contribution, and masked modality autoencoder is introduced to compensate for deficient LiDAR field of view using correlated camera information. These systems demonstrate how distributed learning and intelligent sensing can be co-designed to reinforce each other, enabling IoT systems to collaboratively optimize both data acquisition and model learning under real-world constraints. Taken together, this dissertation establishes a unified system perspective on cooperative edge intelligence through learning and sensing. By bridging communication-efficient federated learning, neural-enhanced edge sensing, and multimodal federated perception, this dissertation provides new system and algorithmic foundations for building scalable, adaptive, and intelligent IoT systems.","Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).","Includes bibliographical references."],"dc:format":["application/pdf","1 online resource (138 p.) : ill."],"dc:identifier":["etd:53026","local:Gan_grad.msu_0128D_22030","https://doi.org/doi:10.25335/kb01-at75"],"dc:language":["eng"],"dc:relation":["Electronic Theses & Dissertations"],"dc:rights":["In Copyright"],"dc:subject":["Computer science"],"dc:title":["TOWARD COOPERATIVE EDGE INTELLIGENCE : INTEGRATING LEARNING AND SENSING IN INTERNET-OF-THINGS SYSTEMS"],"dc:type":["text","Electronic dissertations"]},"updated_at":"2026-11-18T18:59:37Z"}