IEEE Transactions on Network Science and Engineering最新文献_第2页

Protocol-Based Model Predictive Control for Networked Switching Systems With Piecewise-Homogeneous Sojourn Probabilities

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-03-03 DOI: 10.1109/TNSE.2025.3547324

Jun Cheng;Hongjie Pang;Huaicheng Yan;Ju H. Park;Wenhai Qi

{"title":"Protocol-Based Model Predictive Control for Networked Switching Systems With Piecewise-Homogeneous Sojourn Probabilities","authors":"Jun Cheng;Hongjie Pang;Huaicheng Yan;Ju H. Park;Wenhai Qi","doi":"10.1109/TNSE.2025.3547324","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3547324","url":null,"abstract":"Networked switching systems, which integrate multiple subsystems controlled by switching signals, play a crucial role in modern cyber-physical applications such as industrial automation and smart grids. However, their performance is often limited by constrained communication bandwidth and complex dynamic interactions. To address these challenges, this paper proposes a protocol-based model predictive control (MPC) framework for networked switching systems with piecewise-homogeneous sojourn probabilities. A dynamically matching mechanism is designed to quantify mode mismatches caused by network-induced uncertainties. Additionally, an adaptive dynamic-memory event-triggered protocol (ADMETP) is developed, which leverages historical data to optimize triggering decisions and dynamically adjusts thresholds to reduce communication overhead while maintaining system stability. Sufficient conditions for mean-square exponential stability are derived using Lyapunov theory, providing rigorous theoretical guarantees. The effectiveness of the approach is validated through simulations of a numerical experiment and an RLC circuit, demonstrating superior resource utilization and control performance compared to existing methods. This work bridges the gap between adaptive resource management and robust control in networked switching systems, offering practical insights for applications with constrained communication resources.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2322-2332"},"PeriodicalIF":6.7,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870880","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

Exploring the Privacy-Accuracy Trade-Off Using Adaptive Gradient Clipping in Federated Learning

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-28 DOI: 10.1109/TNSE.2025.3546777

Benteng Zhang;Yingchi Mao;Xiaoming He;Ping Ping;Huawei Huang;Jie Wu

{"title":"Exploring the Privacy-Accuracy Trade-Off Using Adaptive Gradient Clipping in Federated Learning","authors":"Benteng Zhang;Yingchi Mao;Xiaoming He;Ping Ping;Huawei Huang;Jie Wu","doi":"10.1109/TNSE.2025.3546777","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3546777","url":null,"abstract":"In Differentially Private Federated Learning (DP-FL), gradient clipping can prevent excessive noise from being added to the gradient and ensure that the impact of noise is within a controllable range. However, state-of-the-art methods adopt fixed or imprecise clipping thresholds for gradient clipping, which is not adaptive to the changes in the gradients. This issue can lead to a significant degradation in accuracy while training the global model. To this end, we propose Differential Privacy Federated Adaptive gradient Clipping based on gradient Norm (DP-FedACN). DP-FedACN can calculate the decay rate of the clipping threshold by considering the overall changing trend of the gradient norm. Furthermore, DP-FedACN can accurately adjust the clipping threshold for each training round according to the actual changes in gradient norm, clipping loss, and decay rate. Experimental results demonstrate that DP-FedACN can maintain privacy protection performance similar to that of DP-FedAvg under member inference attacks and model inversion attacks. DP-FedACN significantly outperforms DP-FedAGNC and DP-FedDDC in privacy protection metrics. Additionally, the test accuracy of DP-FedACN is approximately 2.61%, 1.01%, and 1.03% higher than the other three baseline methods, respectively. DP-FedACN can improve the global model training accuracy while ensuring the privacy protection of the model. All experimental results demonstrate that the proposed DP-FedACN can help find a fine-grained privacy-accuracy trade-off in DP-FL.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2254-2265"},"PeriodicalIF":6.7,"publicationDate":"2025-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870877","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

Efficient and Adaptive CUR Matrix Decomposition for Flexible Compression of Network Monitoring Data

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-28 DOI: 10.1109/TNSE.2025.3546687

Jigang Wen;Shiqin Wang;Kun Xie;Jiazheng Tian;Yixuan Wang

{"title":"Efficient and Adaptive CUR Matrix Decomposition for Flexible Compression of Network Monitoring Data","authors":"Jigang Wen;Shiqin Wang;Kun Xie;Jiazheng Tian;Yixuan Wang","doi":"10.1109/TNSE.2025.3546687","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3546687","url":null,"abstract":"Network-wide monitoring is indispensable for a variety of network applications. However, as network sizes increase and the demand for fine-grained, continuous measurements grows, the challenges associated with storing and transmitting such data intensify. Recent studies have shown that network-wide monitoring data exhibits a low-rank structure, which can be exploited using matrix decomposition techniques for compression. This paper presents a compression algorithm for low-rank matrices based on CUR decomposition, which offers enhanced interpretability compared to SVD-based compression. Existing CUR solutions, however, lack the capability for fast and flexible compression that can dynamically adjust to matrix size requirements while preserving maximal approximation accuracy. We address the challenges associated with CUR row and column selection by formulating it as a deterministic CUR matrix decomposition problem, involving a selection matrix <inline-formula><tex-math>$mathbf{W}$</tex-math></inline-formula>. To achieve rapid compression, we propose an algorithm that effectively accelerates the process of solving for the parameter matrix <inline-formula><tex-math>$mathbf{W}$</tex-math></inline-formula>. Our approach reveals that the vectors in <inline-formula><tex-math>$mathbf{W}$</tex-math></inline-formula> indicate the importance of each row and column in forming the respective row and column subspaces. Leveraging this insight, we develop a flexible compression algorithm based on the sorted vectors in the selection matrix <inline-formula><tex-math>$mathbf{W}$</tex-math></inline-formula>. This method not only ensures the required compression ratio but also maintains maximal approximation accuracy. Extensive experiments on both synthesized and real data demonstrate that our algorithm can deliver fast and precise matrix compression, aligning with the desired compression ratio.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2231-2242"},"PeriodicalIF":6.7,"publicationDate":"2025-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870882","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

Distributed Leader-Following Formation Control of Networked Mobile Robots via Global Orientation Estimation

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-27 DOI: 10.1109/TNSE.2025.3545119

Siqi Wang;Heng Wang;Weiwei Che;Qing Li

{"title":"Distributed Leader-Following Formation Control of Networked Mobile Robots via Global Orientation Estimation","authors":"Siqi Wang;Heng Wang;Weiwei Che;Qing Li","doi":"10.1109/TNSE.2025.3545119","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3545119","url":null,"abstract":"This article proposes a novel distributed leader-following formation control strategy for multiple networked mobile robots, utilizing the relative position measurements among robots. In particular, the leader is assumed to have the same kinematics as the followers and all the robots do not rely on orientation measurements. Firstly, a global orientation estimation law is proposed in the sense that the followers' orientation information is estimated in the leader's reference frame, only based on the leader's orientation estimation and relative bearing information. Secondly, since the leader is not directly connected to all the followers, a new distributed state observer is designed for each follower to estimate the leader's states. Especially, the designed observer not only removes the algebraic loops issue but also eliminates the requirement for the leader's acceleration information. Furthermore, a distributed formation control law is proposed by incorporating the previous estimations, and it is proved that the closed-loop system consisting of the observer and controller is asymptotically stable. Finally, simulation results validate the effectiveness and superiority of the proposed method.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2135-2150"},"PeriodicalIF":6.7,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870955","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

Efficient Federated Learning in Wireless Networks With Incremental Model Quantization and Uploading

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-26 DOI: 10.1109/TNSE.2025.3546333

Zheng Qin;Gang Feng;YiJing Liu;Takshing P. Yum;Fei Wang;Jun Wang

{"title":"Efficient Federated Learning in Wireless Networks With Incremental Model Quantization and Uploading","authors":"Zheng Qin;Gang Feng;YiJing Liu;Takshing P. Yum;Fei Wang;Jun Wang","doi":"10.1109/TNSE.2025.3546333","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3546333","url":null,"abstract":"Federated Learning (FL) has been widely recognized as a promising promoter for future intelligent wireless networks, by collaboratively training a global machine learning (ML) model in a privacy-preserving manner. However, the transmission of large-scale models between clients and servers is susceptible to limited communication resources. Recently proposed model quantization can reduce communication costs by compressing the amount of model data to be transmitted. These methods need to be modified when used in wireless networks with rapidly changing radio channels. In this paper, we propose a federated learning scheme with incremental model quantization and uploading mechanism, called Fed_IQ. Specifically, individual clients quantize the local model parameters to derive the base and incremental model parameters. The base model is first uploaded, while the incremental model is uploaded when the wireless link is sufficiently good. The quantization levels are also adapted to the instantaneous channel states. The server then uses only the base model or combines the base and incremental model to aggregate a more accurate global model. Experimental results show our proposed Fed_IQ can significantly reduce transmission delay and improve model accuracy in a wireless network compared with a number of known state-of-the-art algorithms.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2217-2230"},"PeriodicalIF":6.7,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870879","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

Large-Scale Group Opinion Evolution With Coexistence of Influential Individuals and Strongly Organized Groups Based on Mean Field Games

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-26 DOI: 10.1109/TNSE.2025.3546295

Lu Ren;Yuxin Jin;Wang Yao;Xiao Zhang;Guanghui Jiao

{"title":"Large-Scale Group Opinion Evolution With Coexistence of Influential Individuals and Strongly Organized Groups Based on Mean Field Games","authors":"Lu Ren;Yuxin Jin;Wang Yao;Xiao Zhang;Guanghui Jiao","doi":"10.1109/TNSE.2025.3546295","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3546295","url":null,"abstract":"With the vigorous development of online media, the prevalence of key opinion leaders and water armies has led to unexpected evolutions of users' opinions. Therefore, it is valuable and interesting to investigate the opinion evolution problem for large-scale groups with the coexistence of influential individuals and strongly organized groups. For the above problem, based on the mean-field game theory, this article innovatively proposes a multi-leader multi-population-follower Stackelberg mean field game (MLMPF-SMFG) model to describe the opinion evolution scenario, in which influential individuals are regarded as leaders, normal and strongly organized groups are regarded as follower populations. Moreover, for generality, the types of strongly organized individuals are classified into three typical types: propagandists, spies, and neutrals. Then, the optimal strategies are derived via the adjoint method and solved by forward–backward stochastic differential equations. Sufficient conditions for the existence and uniqueness of the Stackelberg equilibrium (SE) are given, and the approximate SE of the finite system is proven. Finally, simulation experiments on the opinion evolutions of two influential individuals and two ordinary groups are performed to demonstrate the feasibility and effectiveness of the proposed MLMPF-SMFG model.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2202-2216"},"PeriodicalIF":6.7,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143871015","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 Deep Reinforcement Learning for Multimodal Content Caching in Edge-Cloud Networks

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-26 DOI: 10.1109/TNSE.2025.3545924

Weijia Feng;Xinyu Zuo;Ruojia Zhang;Yichen Zhu;Chenyang Wang;Jia Guo;Chuan Sun

{"title":"Federated Deep Reinforcement Learning for Multimodal Content Caching in Edge-Cloud Networks","authors":"Weijia Feng;Xinyu Zuo;Ruojia Zhang;Yichen Zhu;Chenyang Wang;Jia Guo;Chuan Sun","doi":"10.1109/TNSE.2025.3545924","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3545924","url":null,"abstract":"Edge caching presents a promising avenue for mitigating backbone network congestion by strategically caching frequently accessed content at the network periphery. As most current edge caching solutions are designed for single-modal content requests, they cannot deal with the increasing volume of multi-modal content requests. In this study, we investigate the issue of multimodal content caching in edge-cloud networks. Firstly, we establish a heterogeneous edge-cloud network adept at caching multimodal content proximate to end-users to facilitate expeditious content delivery. By leveraging latent representations of multimodal content, we identify distinct user request modalities for multimodal content. Subsequently, we formulate caching replacement operations as a Markov Decision Process (MDP) aimed at minimizing user-content access latency. Moreover, we propose a decentralized multimodal content caching framework at the network edge based on federated deep reinforcement learning. This framework affords distributed decision-making and learning capabilities, thereby alleviating the strain on centralized resources and augmenting caching efficacy. To demonstrate the efficacy of our proposed framework, we conduct comprehensive experiments utilizing the Noah-Wukong dataset. Experimental results provide evidence that our framework reduces average latency by up to 10% compared to traditional methods, highlighting its proficiency in enhancing cache performance in edge-cloud networks.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2188-2201"},"PeriodicalIF":6.7,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870878","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

OD Traffic Maps Recovery for Web 3.0 by Network Tomography in Hankel Tensor Space

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-25 DOI: 10.1109/TNSE.2025.3542603

Chao Wang;Jiuzhen Zeng;Laurence T. Yang;Xiangli Yang;Xianjun Deng;Hao Wang

{"title":"OD Traffic Maps Recovery for Web 3.0 by Network Tomography in Hankel Tensor Space","authors":"Chao Wang;Jiuzhen Zeng;Laurence T. Yang;Xiangli Yang;Xianjun Deng;Hao Wang","doi":"10.1109/TNSE.2025.3542603","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3542603","url":null,"abstract":"In the emerging Web 3.0, origin-destination (OD) traffic maps play a crucial role in network maintenance and management. However, increasing network size and complexity, as well as insufficient or invalid NetFlow protocol-based measurements pose numerous challenges to recovering traffic maps for Web 3.0. This paper therefore proposes RNT-HTT, a robust Network Tomography model based on Hankel time-structured tensor, to accurately recover OD traffic maps with link loads and a fraction of NetFlow counts in Hankel tensor space. More specifically, we propose to Hankelize both OD traffic and link load matrices to three-way tensors along time direction, which fully exploits time-structured correlations concealed in network data. OD pairs-mode product is also designed to model the relation between the Hankelized OD traffic and link load tensors. On the basis of these, RNT-HTT formulates the recovery problem as a convex optimization program with tensor nuclear and <inline-formula><tex-math>${{ell }_{1}}$</tex-math></inline-formula>-norms to respectively effect traffic low-rank and noise sparsity characteristics. In addition, the block-iteration alternating direction method of multipliers (ADMM) and bidirectional pre-sampling schemes are developed to solve RNT-HTT reliably and efficiently. Extensive experiments on three real-world datasets verify effectiveness of RNT-HTT, and corroborate its superior performance over state-of-the-art methods in terms of the recovery accuracy.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"1981-1993"},"PeriodicalIF":6.7,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143871014","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

Mechanism Design for Blockchain Order Books Against Selfish Miners

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-24 DOI: 10.1109/TNSE.2025.3544686

Yunshu Liu;Lingjie Duan

{"title":"Mechanism Design for Blockchain Order Books Against Selfish Miners","authors":"Yunshu Liu;Lingjie Duan","doi":"10.1109/TNSE.2025.3544686","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3544686","url":null,"abstract":"In blockchain-based order book systems, buyers and sellers trade assets, while it is miners to match them and include their transactions in the blockchain. It is found that many miners behave selfishly and myopically, prioritizing transactions with high fees and ignoring many desirable matches that could enhance social welfare. Existing blockchain mechanisms fail to address this issue by overlooking miners' selfish behaviors. To our best knowledge, this work presents the first analytical study to quantify and understand buyer and seller transaction fee choices and selfish miners' transaction matching strategies, proving an infinitely large price of anarchy (PoA) for social welfare loss. To mitigate this, we propose an adjustable block size mechanism that is easy to implement without altering the existing decentralized protocols and still allows buyers and sellers to freely decide transaction fees and miners to selfishly match. The analysis is challenging, as pure strategy Nash equilibria do not always exist, requiring the analysis of many buyers' or sellers' interactive mixed-strategy distributions. Moreover, the system designer may even lack information about each buyer's or seller's bid/ask prices and trading quantities. Nevertheless, our mechanism achieves a well-bounded PoA, and under the homogeneous-quantity trading for non-fungible tokens (NFT), it attains a PoA of 1 with no social welfare loss. We implement our mechanism on a local instance of Ethereum to demonstrate the feasibility of our approach. Experiments based on the realistic dataset demonstrate that our mechanism achieves social optimum for homogeneous-quantity trading like NFT. It can enhance social welfare up to 3.7 times compared to the existing order book benchmarks for heterogeneous-quantity trading of Bitcoin tokens. It exhibits robustness against random variations in buyers and sellers.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2121-2134"},"PeriodicalIF":6.7,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870952","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

Knowledge Efficient Federated Continual Learning for Industrial Edge Systems

IF 6.7 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2025-02-24 DOI: 10.1109/TNSE.2025.3544614

Jiao Chen;Jiayi He;Jianhua Tang;Weihua Li;Zihang Yin

{"title":"Knowledge Efficient Federated Continual Learning for Industrial Edge Systems","authors":"Jiao Chen;Jiayi He;Jianhua Tang;Weihua Li;Zihang Yin","doi":"10.1109/TNSE.2025.3544614","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3544614","url":null,"abstract":"Recent advances in federated learning (FL) primarily focus on addressing inter-client data heterogeneity, implicitly assuming static data within each client. However, this assumption is inadequate for industrial edge systems (IES), which operate in dynamically changing environments and require real-time processing and analysis of voluminous time-series data generated by the Internet of Things. To bridge this gap, we propose MeCo, a novel federated continual learning (FCL) method for IES, designed to avoid forgetting past knowledge while continuously adapting to new task data. MeCo distinguishes itself from traditional FL by effectively addressing both inter-client and intra-client data heterogeneity through a knowledge-efficient strategy. Specifically, it includes: <italic>Meta task-invariant knowledge consolidation,</i> which helps capture shared features across tasks to alleviate forgetting; <italic>Consistent task-specific knowledge transfer,</i> which allows edge clients to extract relevant knowledge from a server-side knowledge pool, providing a jump-starting for the current task. Experimental results demonstrate that MeCo significantly outperforms other federated and/or continual learning approaches in real-world industrial fault diagnosis, achieving approximately 2% higher Mean Average Accuracy and being 1.74 times more cost-effective in server-to-client communication. These advantages, along with its robust performance in IES, indicate the potential of MeCo for facilitating edge-cloud collaborative learning in the future.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 3","pages":"2107-2120"},"PeriodicalIF":6.7,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870957","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