IEEE Transactions on Mobile Computing最新文献_第9页

Communication-Efficient Multi-Server Federated Learning via Over-the-Air Computation 基于无线计算的高效通信多服务器联合学习

IF 9.2 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-26 DOI: 10.1109/TMC.2025.3573600

Rui Han;Jiahao Ma;Lin Bai;Jinho Choi;Wei Zhang

{"title":"Communication-Efficient Multi-Server Federated Learning via Over-the-Air Computation","authors":"Rui Han;Jiahao Ma;Lin Bai;Jinho Choi;Wei Zhang","doi":"10.1109/TMC.2025.3573600","DOIUrl":"https://doi.org/10.1109/TMC.2025.3573600","url":null,"abstract":"Thanks to the Internet of Things (IoT), there has been explosive growth in edge devices, which generate a tremendous amount of data that holds invaluable potential. However, conventional data mining and machine learning (ML) paradigms require transmitting raw data to data centers for further use, which puts a heavy burden on communication networks and is exposed to high privacy risks. Federated learning allows for the training of ML models using distributed datasets, which can be applied to protect data privacy and alleviate transmission burdens. Meanwhile, the technique of over-the-air (OTA) computation can be utilized to exploit the superposition property of wireless communication channels. Motivated by this, in this paper, we propose a co-phase OTA approach for communication-efficient uploading in multi-server federated learning, which does not require expansion of the uplink channel bandwidth when the numbers of users and models increase. Besides, the digital OTA with randomized transmission is proposed to overcome the disadvantages of analog OTA, where the performance analyses of analog OTA and digital OTA are deduced, respectively. Simulation results show that a lower cost function can be obtained by digital OTA while requiring fewer iterations for convergence than that in analog OTA as more users can upload.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 10","pages":"10683-10695"},"PeriodicalIF":9.2,"publicationDate":"2025-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145036758","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Joint Trajectory and Beamforming Optimization for AAV-Relayed Integrated Sensing and Communication With Mobile Edge Computing 基于移动边缘计算的aav中继集成传感与通信联合轨迹和波束成形优化

IF 9.2 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-26 DOI: 10.1109/TMC.2025.3573702

Shanfeng Xu;Zhipeng Liu;Le Zhao;Ziyi Liu;Xinyi Wang;Zesong Fei;Arumugam Nallanathan

{"title":"Joint Trajectory and Beamforming Optimization for AAV-Relayed Integrated Sensing and Communication With Mobile Edge Computing","authors":"Shanfeng Xu;Zhipeng Liu;Le Zhao;Ziyi Liu;Xinyi Wang;Zesong Fei;Arumugam Nallanathan","doi":"10.1109/TMC.2025.3573702","DOIUrl":"https://doi.org/10.1109/TMC.2025.3573702","url":null,"abstract":"In this paper, we investigate joint trajectory and beamforming design for autonomous aerial vehicle (AAV)-relayed integrated sensing and communication (ISAC) systems with mobile edge eomputing (MEC) under the clutter environment. Due to the limited on-board computing capability, the AAV has to offload sensing echoes to the base station (BS) for efficient processing. A novel relay-based ISAC-then-offload frame structure is considered. We aim to maximize the throughput of the BS-AAV-user relaying link while ensuring sensing accuracy and efficient sensing data offloading. The non-convex problem is solved using an alternating optimization algorithm based on successive convex approximation (SCA). Simulation results illustrate that our proposed algorithm achieves near-optimal communication performance while guaranteeing sensing accuracy, addressing the balance between the communication and sensing performance. Furthermore, we evaluate the impact of critical system parameters including sensing constraints, power control factor, and AAV flight duration on communication performance, and explore the trade-offs between energy efficiency and spectral efficiency under varying sensing data intensity and offloading duration.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 10","pages":"11180-11192"},"PeriodicalIF":9.2,"publicationDate":"2025-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145021380","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Sequential Federated Learning in Hierarchical Architecture on Non-IID Datasets 非iid数据集上层次结构中的顺序联邦学习

IF 9.2 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-26 DOI: 10.1109/TMC.2025.3573928

Xingrun Yan;Shiyuan Zuo;Rongfei Fan;Han Hu;Li Shen;Puning Zhao;Yong Luo

{"title":"Sequential Federated Learning in Hierarchical Architecture on Non-IID Datasets","authors":"Xingrun Yan;Shiyuan Zuo;Rongfei Fan;Han Hu;Li Shen;Puning Zhao;Yong Luo","doi":"10.1109/TMC.2025.3573928","DOIUrl":"https://doi.org/10.1109/TMC.2025.3573928","url":null,"abstract":"In a real federated learning (FL) system, communication overhead for passing model parameters between the clients and the parameter server (PS) is often a bottleneck. Hierarchical federated learning (HFL) that poses multiple edge servers (ESs) between clients and the PS can partially alleviate communication pressure but still needs the aggregation of model parameters from multiple ESs at the PS. To further reduce communication overhead, we remove the central PS, so that each iteration only completes model training by transmitting the global model between two adjacent ES. We call this serial learning method Sequential FL (SFL). For the first time, we introduced SFL into HFL and proposed a novel algorithm adapted to this combined framework, called Fed-CHS. Convergence results are derived for strongly convex and non-convex loss functions under various data heterogeneity setups, which show comparable convergence performance with the algorithms for HFL or SFL solely. Experimental results provide evidence of the superiority of our proposed Fed-CHS on both communication overhead saving and test accuracy over baseline methods.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 10","pages":"11110-11124"},"PeriodicalIF":9.2,"publicationDate":"2025-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145021223","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Budget-Feasible Diffusion Mechanisms for Mobile Crowdsourcing in Social Networks 社交网络中移动众包的预算可行扩散机制

IF 7.7 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-24 DOI: 10.1109/TMC.2025.3549751

Xiang Liu;Weiwei Wu;Minming Li;Wanyuan Wang;Yifan Qin;Yingchao Zhao;Junzhou Luo

{"title":"Budget-Feasible Diffusion Mechanisms for Mobile Crowdsourcing in Social Networks","authors":"Xiang Liu;Weiwei Wu;Minming Li;Wanyuan Wang;Yifan Qin;Yingchao Zhao;Junzhou Luo","doi":"10.1109/TMC.2025.3549751","DOIUrl":"https://doi.org/10.1109/TMC.2025.3549751","url":null,"abstract":"Mobile crowdsourcing has emerged as a popular approach for organizations to leverage the collective intelligence of a crowd of users to obtain services. Considering users’ costs for providing services, it is vital for the requester to design incentive mechanisms to encourage users’ participation in crowdsourcing under the budget constraint. This aligns with the concept of budget-feasible mechanism design. Existing budget-feasible mechanisms often assume immediate user reachability and willingness of joining the crowdsourcing, which is unrealistic. To address this issue, a promising approach is to have participating users diffuse auction information to potential users in the social network. However, this brings another challenge in that participating users can be strategic and therefore hesitant to invite more potential competitors to join the crowdsourcing platform. In this paper, we focus on developing diffusion mechanisms that incentivize strategic users to actively diffuse auction information through the social network. This helps to attract more informed users and ultimately increases the value of the procured services. Specifically, we propose optimal budget-feasible diffusion mechanisms that simultaneously guarantee individual rationality, budget-feasibility, strong budget-balance, incentive-compatibility (i.e., users report real costs and diffuse auction information to all their neighbors) and approximation. Experiment results under real datasets further demonstrate the efficiency of proposed mechanisms.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 8","pages":"7189-7205"},"PeriodicalIF":7.7,"publicationDate":"2025-03-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144550536","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

AIChronoLens: AI/ML Explainability for Time Series Forecasting in Mobile Networks AIChronoLens：移动网络中时间序列预测的AI/ML可解释性

IF 7.7 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-24 DOI: 10.1109/TMC.2025.3554035

Pablo Fernández Pérez;Claudio Fiandrino;Eloy Pérez Gómez;Hossein Mohammadalizadeh;Marco Fiore;Joerg Widmer

{"title":"AIChronoLens: AI/ML Explainability for Time Series Forecasting in Mobile Networks","authors":"Pablo Fernández Pérez;Claudio Fiandrino;Eloy Pérez Gómez;Hossein Mohammadalizadeh;Marco Fiore;Joerg Widmer","doi":"10.1109/TMC.2025.3554035","DOIUrl":"https://doi.org/10.1109/TMC.2025.3554035","url":null,"abstract":"Forecasting is increasingly considered a fundamental enabler for the management of next-generation mobile networks. While deep neural networks excel at short- and long-term forecasting, their complexity hinders interpretability, a crucial factor for production deployment. The existing EXplainable Artificial Intelligence (XAI) techniques, primarily designed for computer vision and natural language processing, struggle with time series data due to their lack of understanding of temporal characteristics of the input data. In this paper, we take the research on EXplainable Artificial Intelligence (XAI) for time series forecasting one step further by proposing <sc>AIChronoLens</small>, a new tool that links legacy XAI explanations with the temporal properties of the input. <sc>AIChronoLens</small> allows diving deep into the behavior of time series predictors and spotting, among other aspects, the hidden causes of forecast errors. We show that <sc>AIChronoLens</small>’s output can be utilized for meta-learning to predict when the original time series forecasting model makes errors and fix them in advance, thereby improving the accuracy of the predictors. Extensive evaluations with real-world mobile traffic traces pinpoint model behaviors that would not be possible to identify otherwise and show how model performance can be improved by 32 % upon re-training and by up to 39 % with meta-learning.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 8","pages":"7757-7772"},"PeriodicalIF":7.7,"publicationDate":"2025-03-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144550755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

mmZeAR: Zero-Effort Cross-Category Action Recognition With mmWave Radar mmZeAR：毫米波雷达零努力跨类别动作识别

IF 9.2 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-23 DOI: 10.1109/TMC.2025.3573168

Biyun Sheng;Jiabin Li;Hui Cai;Yiping Zuo;Li Lu;Fu Xiao

{"title":"mmZeAR: Zero-Effort Cross-Category Action Recognition With mmWave Radar","authors":"Biyun Sheng;Jiabin Li;Hui Cai;Yiping Zuo;Li Lu;Fu Xiao","doi":"10.1109/TMC.2025.3573168","DOIUrl":"https://doi.org/10.1109/TMC.2025.3573168","url":null,"abstract":"Despite the widespread application of radio frequency (RF) signal-based human action recognition, traditional solutions can only recognize seen categories and the perception scope is restrained by the limited activity classes. When a novel category emerges, the model needs to be optimized again on additionally collected samples at the cost of computation and labor burden. To address this challenge, we develop the mmZeAR system, which learns semantic knowledge from available vision data as class attributes and then transforms the classification into a matching problem. Specifically, we build the attribute space by fusing the coarse-grained video classification features and fine-grained angle change features of 3D joint skeletons. Then we design an efficient feature extraction backbone named TriSqN, which integrates triple radar heatmaps into the final representations by sufficiently exploring the heterogeneous and complementary characteristics. Finally, a projection network is developed between semantic attributes and radar features to construct indirect relationships between samples and labels. By implementing mmZeAR on millimeter wave (mmWave) radar signal datasets, our extensive experiments have demonstrated its remarkable recognition accuracy in novel category recognition with zero effort and achieved state-of-the-art performance.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 10","pages":"11164-11179"},"PeriodicalIF":9.2,"publicationDate":"2025-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145021302","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multi-Source Domain Generalization for CSI-Based Human Activity Recognition 基于csi的人类活动识别多源域概化

IF 9.2 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-23 DOI: 10.1109/TMC.2025.3573457

Tianqi Fan;Sen Qiu;Wei Gong;Yuguang Fang

{"title":"Multi-Source Domain Generalization for CSI-Based Human Activity Recognition","authors":"Tianqi Fan;Sen Qiu;Wei Gong;Yuguang Fang","doi":"10.1109/TMC.2025.3573457","DOIUrl":"https://doi.org/10.1109/TMC.2025.3573457","url":null,"abstract":"Domain generalization remains a key challenge in human activity recognition based on channel state information (CSI). Different domains correspond to distinct data distributions, deviating from the typical assumption of independent and identically distributed (i.i.d.) data, which leads to significant performance degradation when models are applied to unseen domains. To address this issue, we propose a novel domain generalization model that integrates meta-learning initialization and an adaptive channel grouping attention mechanism. First, a meta-learning strategy is employed to acquire well-initialized parameters from multiple source domain tasks, enabling the model to implicitly enhance its cross-domain generalization ability. Second, an adaptive grouping attention mechanism is designed in the feature extraction stage to effectively capture the sensitivity differences of different subcarriers to human activities. Meanwhile, a random masking training mechanism is introduced to simulate real-world domain variations and improve model robustness. In addition, a domain adversarial training framework based on the gradient reversal layer (GRL) is adopted to mitigate domain-specific feature dependency, further enhancing the model’s generalization capability. We evaluate our proposed method on both a self-collected dataset, which includes human activity data from nine volunteers across six different environments, and a public CSI dataset. The experimental results demonstrate that our method significantly outperforms existing approaches in domain generalization performance, verifying its effectiveness and practical applicability.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 10","pages":"11034-11045"},"PeriodicalIF":9.2,"publicationDate":"2025-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145021366","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Dual Dependency-Aware Collaborative Service Caching and Task Offloading in Vehicular Edge Computing 车辆边缘计算中双依赖感知协同服务缓存和任务卸载

IF 9.2 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-23 DOI: 10.1109/TMC.2025.3573379

Liang Zhao;Lu Sun;Ammar Hawbani;Zhi Liu;Xiongyan Tang;Lexi Xu

{"title":"Dual Dependency-Aware Collaborative Service Caching and Task Offloading in Vehicular Edge Computing","authors":"Liang Zhao;Lu Sun;Ammar Hawbani;Zhi Liu;Xiongyan Tang;Lexi Xu","doi":"10.1109/TMC.2025.3573379","DOIUrl":"https://doi.org/10.1109/TMC.2025.3573379","url":null,"abstract":"Although some studies in recent years have focused on the coexistence of service and task dependencies in the collaborative optimization of service caching and task offloading in Vehicle Edge Computing (VEC), the challenges brought by dual dependencies have not been fully addressed. Therefore, this paper proposes a more comprehensive joint optimization method for service caching and task offloading under dual dependencies. First, this paper proposes a service criticality prediction method based on the Gated Graph Recurrent Network (GGRN) to perceive complex task dependencies and accurately capture the service requirements of critical task types. Based on this, a hierarchical active-passive hybrid caching strategy is designed, which aims to satisfy diverse service demands while reducing the additional overhead caused by remote service requests. Second, a global task priority computation method based on application heterogeneity has been developed to prevent cascading delays in task chains. Finally, this paper formulates a joint optimization problem for service caching and task offloading in a three-layer VEC system, models it as a markov decision process, and applies a proximal policy optimization-driven collaborative optimization algorithm named COHCTO. Simulation results show that COHCTO achieves multi-objective optimization across metrics such as delay, energy consumption, caching hit rate, and application success rate under conditions different from those of other algorithms.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 10","pages":"10963-10977"},"PeriodicalIF":9.2,"publicationDate":"2025-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145021388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Toward Generalized Urban Computing: Pretraining a Spatial-Temporal Model for Diverse Urban Tasks 面向广义城市计算：不同城市任务的时空模型预训练

IF 9.2 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-23 DOI: 10.1109/TMC.2025.3573373

Yingqian Zhang;Chao Li;Shibo He;Xiangliang Zhang;Jiming Chen

{"title":"Toward Generalized Urban Computing: Pretraining a Spatial-Temporal Model for Diverse Urban Tasks","authors":"Yingqian Zhang;Chao Li;Shibo He;Xiangliang Zhang;Jiming Chen","doi":"10.1109/TMC.2025.3573373","DOIUrl":"https://doi.org/10.1109/TMC.2025.3573373","url":null,"abstract":"Urban computing leverages data analysis to improve urban areas’ efficiency and sustainability, tackling tasks like traffic management, crime forecasting, and air quality predictions. Current models, while efficient, often struggle with tasks beyond their initial training due to limited flexibility. Typically, new tasks require developing specialized models, which may not perform optimally with limited data. To overcome these challenges, we propose the development of a universal pretrained model that understands a city’s various aspects comprehensively. This model serves as a robust foundation, ready to be quickly adjusted for different urban tasks as they arise, even if they occur in different cities. Unlike language models, urban computing models must handle unique spatial-temporal dynamics, making standard pretraining techniques inadequate. Our approach includes a spatial-temporal module with multi-graph convolution and temporal attention mechanisms, capturing the necessary spatial-temporal patterns during pretraining. We also integrate a prompt-tuning module within this framework, which can be adapted for new predictive tasks. The results of extensive experiments on four urban predictive tasks across two cities demonstrate the effectiveness of our model.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 10","pages":"10840-10852"},"PeriodicalIF":9.2,"publicationDate":"2025-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145036742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Federated Meta-Learning Based Computation Offloading Approach With Energy-Delay Tradeoffs in UAV-Assisted VEC 基于联邦元学习的无人机辅助VEC能量延迟权衡计算卸载方法

IF 9.2 2区计算机科学

IEEE Transactions on Mobile Computing Pub Date : 2025-03-23 DOI: 10.1109/TMC.2025.3573278

Chunlin Li;Chaoyue Deng;Yong Zhang;Shaohua Wan

{"title":"Federated Meta-Learning Based Computation Offloading Approach With Energy-Delay Tradeoffs in UAV-Assisted VEC","authors":"Chunlin Li;Chaoyue Deng;Yong Zhang;Shaohua Wan","doi":"10.1109/TMC.2025.3573278","DOIUrl":"https://doi.org/10.1109/TMC.2025.3573278","url":null,"abstract":"Federated learning (FL) provides an applicable solution for computation offloading in Unmanned Aerial Vehicle(UAV)-assisted Vehicular Edge Computing (VEC) by preserving privacy. However, the heterogeneity of clients brings challenges to the generalization of models. Therefore, we propose a federated meta-learning (FML) framework to solve computation offloading for UAV-assisted VEC. In this paper, we are concerned with computation offloading of temporary hotspot regions due to traffic congestion. First, we construct a computation offloading problem with energy-delay tradeoffs and convert the problem to a Markov Decision Process (MDP). Then, we use FML to train personalized models for different vehicles while enhancing the generalization, we propose a Graph neural network-based FL Probabilistic Embedding for Actor-critic RL (GFL-PEARL) algorithm. We model the context as a Directed Acyclic Graph (DAG) and use GNN to reconstruct the inference network of the PEARL algorithm to extract the correlation between contexts fully. We dynamically adjust the task priority during the FML training process to improve the sampling efficiency. Finally, we verify the performance of the algorithm through simulation and physical experiments. Experimental results show that our algorithm can reduce average cost and task overtime rate by 31% and 56% respectively compared with the benchmarks.","PeriodicalId":50389,"journal":{"name":"IEEE Transactions on Mobile Computing","volume":"24 10","pages":"10978-10991"},"PeriodicalIF":9.2,"publicationDate":"2025-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145021409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0