IEEE Transactions on Signal and Information Processing over Networks最新文献_第7页

R-FAST: Robust Fully-Asynchronous Stochastic Gradient Tracking Over General Topology R-FAST：通用拓扑上的鲁棒全同步随机梯度跟踪

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-08-30 DOI: 10.1109/TSIPN.2024.3444484

Zehan Zhu;Ye Tian;Yan Huang;Jinming Xu;Shibo He

{"title":"R-FAST: Robust Fully-Asynchronous Stochastic Gradient Tracking Over General Topology","authors":"Zehan Zhu;Ye Tian;Yan Huang;Jinming Xu;Shibo He","doi":"10.1109/TSIPN.2024.3444484","DOIUrl":"https://doi.org/10.1109/TSIPN.2024.3444484","url":null,"abstract":"We propose a Robust Fully-Asynchronous Stochastic Gradient Tracking method (R-FAST) for distributed machine learning problems over a network of nodes, where each node performs local computation and communication at its own pace without any form of synchronization. Different from existing asynchronous distributed algorithms, R-FAST can eliminate the impact of data heterogeneity across nodes on convergence performance and allow for packet losses by employing a robust gradient tracking strategy that relies on properly designed auxiliary variables for tracking and buffering the overall gradient vector. Moreover, the proposed method utilizes two spanning-tree graphs for communication so long as both share at least one common root, enabling flexible designs in communication topologies. We show that R-FAST converges in expectation to a neighborhood of the optimum with a geometric rate for smooth and strongly convex objectives; and to a stationary point with a sublinear rate for general non-convex problems. Extensive experiments demonstrate that R-FAST runs 1.5-2 times faster than synchronous benchmark algorithms, such as Ring-AllReduce and D-PSGD, while still achieving comparable accuracy, and outperforms the existing well-known asynchronous algorithms, such as AD-PSGD and OSGP, especially in the presence of stragglers.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"10 ","pages":"665-678"},"PeriodicalIF":3.0,"publicationDate":"2024-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142173995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 Nash Equilibrium Seeking for Nonlinear Players With Input Delay 有输入延迟的非线性参与者的分布式纳什均衡寻求

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-08-29 DOI: 10.1109/TSIPN.2024.3451979

Zhaoming Sheng;Qian Ma

引用次数: 0

Stable Outlier-Robust Signal Recovery Over Networks: A Convex Analytic Approach Using Minimax Concave Loss 网络上稳定的离群稳健信号恢复：使用最小凹损失的凸分析方法

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-08-29 DOI: 10.1109/TSIPN.2024.3451992

Maximilian H. V. Tillmann;Masahiro Yukawa

引用次数: 0

Barycentric Coordinate-Based Distributed Localization Over 3-D Underwater Wireless Sensor Networks 基于重心坐标的三维水下无线传感器网络分布式定位

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-08-08 DOI: 10.1109/TSIPN.2024.3440644

Lei Shi;Shaojie Yao;Nianwen Ning;Yi Zhou

引用次数: 0

Causal Learning and Knowledge Fusion Mechanism for Brain Functional Network Classification 大脑功能网络分类的因果学习和知识融合机制

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-07-18 DOI: 10.1109/TSIPN.2024.3430474

Junzhong Ji;Feipeng Wang;Lu Han;Jinduo Liu

{"title":"Causal Learning and Knowledge Fusion Mechanism for Brain Functional Network Classification","authors":"Junzhong Ji;Feipeng Wang;Lu Han;Jinduo Liu","doi":"10.1109/TSIPN.2024.3430474","DOIUrl":"10.1109/TSIPN.2024.3430474","url":null,"abstract":"Current studies have shown that the classification of human brain functional networks (BFN) is a reliable way to diagnose and predict brain diseases. However, a great challenge for current traditional machine learning methods and deep learning methods is their poor performance or lack of interpretability. To alleviate this problem, we propose a novel causal learning and knowledge fusion mechanism for brain functional network classification, named CLKF. The proposed mechanism first extracts causal relationships among brain regions from functional magnetic resonance imaging (fMRI) data using partial correlation and conditional mutual information, and obtains the relationships between BFN and labels by Gaussian kernel density estimation. Then, it fuses these two types of relationships as knowledge to aid in the classification of brain functional networks. The experimental results on the simulated resting-state fMRI dataset show that the proposed mechanism can effectively learn the causal relationships among brain regions. The results on the real resting-state fMRI dataset demonstrate that our mechanism can not only improve the classification performance of both traditional machine learning and deep learning methods but also provide an interpretation of the results obtained by deep learning methods. These findings suggest that the proposed mechanism has good potential in practical medical applications.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"10 ","pages":"650-664"},"PeriodicalIF":3.0,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141745272","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 Constrained Optimization Algorithm for Higher-Order Multi-Agent Systems 高阶多代理系统的分布式约束优化算法

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-07-18 DOI: 10.1109/TSIPN.2024.3430492

Xiasheng Shi;Lingfei Su;Qing Wang

{"title":"Distributed Constrained Optimization Algorithm for Higher-Order Multi-Agent Systems","authors":"Xiasheng Shi;Lingfei Su;Qing Wang","doi":"10.1109/TSIPN.2024.3430492","DOIUrl":"10.1109/TSIPN.2024.3430492","url":null,"abstract":"The distributed nonsmooth constrained optimization problems over higher-order systems are investigated in this study. The challenges lies in the fact that the output of the agent is directly controlled by the state variable rather than the control input. Compared to existing works, the local objective function is merely assumed to be nonsmooth. Firstly, an initialization-free fully distributed derivative feedback control scheme is developed for the known objective function over double-integrator systems. The local generic constraint is addressed by an adaptive nonnegative penalty factor. Secondly, an initialization-free fully distributed state feedback control scheme is proposed for the unknown objective function over double-integrator systems. Addressing the local box constraint involves incorporating an adaptive penalty factor. Thirdly, the above two algorithms are extended to the general higher-order systems using the tracking control method. In addition, the above-developed methods are proved to be asymptotically convergent under certain conditions. Eventually, the efficiency of the above-produced methods is shown via four simulation cases.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"10 ","pages":"626-639"},"PeriodicalIF":3.0,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141745314","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 Convex Optimization “Over-the-Air” in Dynamic Environments 动态环境中的分布式 "空中 "凸优化

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-07-05 DOI: 10.1109/TSIPN.2024.3423668

Navneet Agrawal;Renato Luís Garrido Cavalcante;Masahiro Yukawa;Sławomir Stańczak

{"title":"Distributed Convex Optimization “Over-the-Air” in Dynamic Environments","authors":"Navneet Agrawal;Renato Luís Garrido Cavalcante;Masahiro Yukawa;Sławomir Stańczak","doi":"10.1109/TSIPN.2024.3423668","DOIUrl":"10.1109/TSIPN.2024.3423668","url":null,"abstract":"This paper presents a decentralized algorithm for solving distributed convex optimization problems in dynamic networks with time-varying objectives. The unique feature of the algorithm lies in its ability to accommodate a wide range of communication systems, including previously unsupported ones, by abstractly modeling the information exchange in the network. Specifically, it supports a novel communication protocol based on the “over-the-air” function computation (OTA-C) technology, that is designed for an efficient and truly decentralized implementation of the consensus step of the algorithm. Unlike existing OTA-C protocols, the proposed protocol does not require the knowledge of network graph structure or channel state information, making it particularly suitable for decentralized implementation over ultra-dense wireless networks with time-varying topologies and fading channels. Furthermore, the proposed algorithm synergizes with the “superiorization” methodology, allowing the development of new distributed algorithms with enhanced performance for the intended applications. The theoretical analysis establishes sufficient conditions for almost sure convergence of the algorithm to a common time-invariant solution for all agents, assuming such a solution exists. Our algorithm is applied to a real-world distributed random field estimation problem, showcasing its efficacy in terms of convergence speed, scalability, and spectral efficiency. Furthermore, we present a superiorized version of our algorithm that achieves faster convergence with significantly reduced energy consumption compared to the unsuperiorized algorithm.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"10 ","pages":"610-625"},"PeriodicalIF":3.0,"publicationDate":"2024-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141577659","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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

Memory-Based Accelerated Consensus for Second-Order Multi-Agent Systems With Delay 基于内存的延迟二阶多代理系统加速共识

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-07-04 DOI: 10.1109/TSIPN.2024.3417225

Mengya Huang;Jing-Wen Yi;Li Chai

引用次数: 0

Distributed Control in Uncertain Nonlinear Multiagent Systems Under Event-Triggered Communication and General Directed Graphs 事件触发通信和一般有向图条件下不确定非线性多代理系统的分布式控制

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-07-03 DOI: 10.1109/TSIPN.2024.3422878

Gang Wang;Zongyu Zuo;Peng Li

{"title":"Distributed Control in Uncertain Nonlinear Multiagent Systems Under Event-Triggered Communication and General Directed Graphs","authors":"Gang Wang;Zongyu Zuo;Peng Li","doi":"10.1109/TSIPN.2024.3422878","DOIUrl":"10.1109/TSIPN.2024.3422878","url":null,"abstract":"Designing a consensus algorithm for multiagent systems within an event-triggered communication setting is challenging due to the discontinuous and inaccurate interaction information caused by event-triggering mechanisms. Currently, most related results require an undirected or balanced directed graph. To avoid such restrictive requirements and consider general directed graphs with a spanning tree, we first investigate the perturbed consensus problem of first-order dynamics. Then, we extend our findings to address the consensus problem of the uncertain nonlinear multiagent systems described in Lagrangian dynamics, Brunovsky dynamics, and strict-feedback dynamics under event-triggered communication. We develop three distributed consensus protocols that consider the unique characteristics of these systems and assign different reference signals accordingly. Our proposed schemes ensure that consensus errors either converge to zero or to a small adjustable neighborhood around zero without Zeno behavior while preserving signal boundedness in the closed-loop system. Finally, we conduct extensive simulations to further illustrate the efficiency of our theoretical results.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"10 ","pages":"599-609"},"PeriodicalIF":3.0,"publicationDate":"2024-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141547847","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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

Fully Distributed Model-Free Flocking of Multiple Euler-Lagrange Systems 多个欧拉-拉格朗日系统的全分布式无模型成群飞行

IF 3 3区计算机科学

IEEE Transactions on Signal and Information Processing over Networks Pub Date : 2024-06-26 DOI: 10.1109/TSIPN.2024.3419437

Mingkang Long;Yin Chen;Housheng Su

引用次数: 0