IEEE Transactions on Signal and Information Processing over Networks最新文献

{"title":"Learning Networks From Wide-Sense Stationary Stochastic Processes","authors":"Anirudh Rayas;Jiajun Cheng;Rajasekhar Anguluri;Deepjyoti Deka;Gautam Dasarathy","doi":"10.1109/TSIPN.2025.3583488","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3583488","url":null,"abstract":"Complex networked systems driven by latent inputs are common in fields like neuroscience, finance, and engineering. A key inference problem here is to learn edge connectivity from node outputs (potentials). We focus on systems governed by steady-state linear conservation laws: <inline-formula><tex-math>$X_{t} = {L^{ast }}Y_{t}$</tex-math></inline-formula>, where <inline-formula><tex-math>$X_{t}, Y_{t} in mathbb {R}^{p}$</tex-math></inline-formula> denote inputs and potentials, respectively, and the sparsity pattern of the <inline-formula><tex-math>$p times p$</tex-math></inline-formula> Laplacian <inline-formula><tex-math>$L^{ast }$</tex-math></inline-formula> encodes the edge structure. Assuming <inline-formula><tex-math>$X_{t}$</tex-math></inline-formula> to be a wide-sense stationary stochastic process with a known spectral density matrix, we learn the support of <inline-formula><tex-math>$L^{ast }$</tex-math></inline-formula> from temporally correlated samples of <inline-formula><tex-math>$Y_{t}$</tex-math></inline-formula> via an <inline-formula><tex-math>$ell _{1}$</tex-math></inline-formula>-regularized Whittle’s maximum likelihood estimator (MLE). The regularization is particularly useful for learning large-scale networks in the high-dimensional setting where the network size <inline-formula><tex-math>$p$</tex-math></inline-formula> significantly exceeds the number of samples <inline-formula><tex-math>$n$</tex-math></inline-formula>. We show that the MLE problem is strictly convex, admitting a unique solution. Under a novel mutual incoherence condition and certain sufficient conditions on <inline-formula><tex-math>$(n, p, d)$</tex-math></inline-formula>, we show that the ML estimate recovers the sparsity pattern of <inline-formula><tex-math>$L^ast$</tex-math></inline-formula> with high probability, where <inline-formula><tex-math>$d$</tex-math></inline-formula> is the maximum degree of the graph underlying <inline-formula><tex-math>$L^{ast }$</tex-math></inline-formula>. We provide recovery guarantees for <inline-formula><tex-math>$L^ast$</tex-math></inline-formula> in element-wise maximum, Frobenius, and operator norms. Finally, we complement our theoretical results with several simulation studies on synthetic and benchmark datasets, including engineered systems (power and water networks), and real-world datasets from neural systems (such as the human brain).","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"655-669"},"PeriodicalIF":3.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144634609","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}

{"title":"GGAT: Gravitation-Based Graph Attention Networks","authors":"Shujuan Wei;Huijun Tang;Pengfei Jiao;Huaming Wu","doi":"10.1109/TSIPN.2025.3583355","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3583355","url":null,"abstract":"Graph-structured data is an important data form that is widely used in the real world. It can effectively and abstractly express entities in information and the relationships between entities. The appearance of Graph Neural Networks (GNNs) provides a potent tool for dealing with nonlinear data structures, which mainly learns node representation through information propagation and aggregation on the nodes in the graph. However, existing GNNs fail to adequately and efficiently integrate the topological structure of the network and node features during information propagation, resulting in an insufficient capture of the complex influence relationships between nodes. The limitation constrains the expression ability of the models and seriously impacts their performance in node classification tasks. To overcome this issue, we propose a Gravitation-based Graph Attention Network (GGAT) for node classification. Firstly, we define a novel similarity measurement method based on the formula of universal gravitation, which combines node information entropy and spatial distance. This method overcomes the limitation of existing similarity measurements that focus solely on the topological structure or node features, achieving a more comprehensive similarity assessment. Then, we apply it to the graph attention network as a novel attention mechanism. Compared with the traditional attention mechanisms based on learning, our proposed mechanism not only thoroughly considers the topological structure and node features to allocate the weights of neighbor nodes but also makes the calculation of attention weights more transparent with an intuitive physical significance, thereby improving the stability and interpretability of the model. Finally, the experiments are carried out on various real datasets, and the results show that GGAT is superior to the existing popular models in node classification performance.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"630-640"},"PeriodicalIF":3.0,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144581557","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}

{"title":"Online Non-Convex Non-Cooperative Cluster-Based Games With Byzantine Resiliency in Decentralized Multi-Agent Systems","authors":"Olusola T. Odeyomi;Temitayo O. Olowu;Opeyemi Ajibuwa;Abdollah Homaifar","doi":"10.1109/TSIPN.2025.3579245","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3579245","url":null,"abstract":"Decentralized multi-agent systems are well known for their ability to model complex systems, such as smart grids, autonomous vehicles, etc. Many decentralized multi-agent systems can be modeled as cluster-based non-cooperative games in which agents within a cluster have selfish interests different from those of agents in other clusters. In this paper, we consider a cluster-based non-cooperative game for multi-agent systems in the presence of Byzantine attacks. This is an area of research yet to be explored in non-cooperative games. Therefore, we propose a novel Byzantine-resilient online mirror descent-based decentralized Nash algorithm. We assume that the loss function is time-varying and non-convex. Also, the agents within each cluster form an unbalanced graph network. Our theoretical and simulation results show that the proposed algorithm is resilient against Byzantine attacks and computationally efficient.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"752-766"},"PeriodicalIF":3.0,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144687641","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}

{"title":"Adaptive Event-Triggered Output Synchronization of Heterogeneous Multiagent Systems: A Model-Free Reinforcement Learning Approach","authors":"Wenfeng Hu;Xuan Wang;Meichen Guo;Biao Luo;Tingwen Huang","doi":"10.1109/TSIPN.2025.3578759","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3578759","url":null,"abstract":"This paper proposes a reinforcement learning approach to the output synchronization problem for heterogeneous leader-follower multi-agent systems, where the system dynamics of all agents are completely unknown. First, to solve the challenge caused by unknown dynamics of the leader, we develop an experience-replay learning method to estimate the leader’s dynamics, which only uses the leader’s past state and output information as training data. Second, based on the newly estimated leader’s dynamics, we design an event-triggered observer for each follower to estimate the leader’s state and output. Furthermore, the experience-replay learning method and the event-triggered leader observer are co-designed, which ensures the convergence and Zeno behavior exclusion. Subsequently, to free the followers from reliance on system dynamics, a data-driven adaptive dynamic programming (ADP) method is presented to iteratively derive the optimal control gains, based on which we design a policy iteration (PI) algorithm for output synchronization. Finally, the proposed algorithm’s performance is validated through a simulation.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"604-616"},"PeriodicalIF":3.0,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144524395","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}

{"title":"State Estimation for Discrete-Time Complex Networks With Time-Varying Outer Coupling and Uncertain Inner Coupling: A Distributed Method","authors":"Chulin Zhou;Shiyou Chen;Hao Liu","doi":"10.1109/TSIPN.2025.3578778","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3578778","url":null,"abstract":"A distributed filtering issue is considered for discrete-time complex networks (CNs) with variable outer coupling and uncertain inner coupling. The outer coupling is described by a matrix with time-varying configuration parameters, and the inner coupling parameters are permitted to change within specific limits. The focus of this study is the design of filters for each node utilizing data from local and neighboring nodes. We prove the estimation error covariance (EEC) is exponentially bounded on the mean square, obtaining an optimal filter gain such that the bound is minimized. And we quantitatively analyzed the relationship between coupling parameters and filtering performance. A series of numerical simulations are given with comparisons to illustrate the filtering performance.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"592-603"},"PeriodicalIF":3.0,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144550760","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}

{"title":"Event-Triggered Adaptive Tracking Control for Multi-Agent Systems With Multiple Uncertainties","authors":"Yong Xu;Meng-Ying Wan;Chong-Yang Wei;Zheng-Guang Wu","doi":"10.1109/TSIPN.2025.3572728","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3572728","url":null,"abstract":"The existing distributed tracking control for heterogeneous multi-agent systems employs a method that involves designing distributed observers relying on precise models. When the system matrices, controller gains, and coupling parameters are unknown, existing control methods struggle to handle multiple unknown parameters concurrently. To address this challenge, we propose a direct adaptive control approach featuring the simplest discrete communication and control structure for studying the event-triggered tracking control problem in heterogeneous and uncertain multi-agent systems. Firstly, we establish two fundamental lemmas pertinent to event-triggered distributed tracking control. Subsequently, we propose a new adaptive event-triggered control strategy featuring the simplest communication architecture, grounded in the two fundamental lemmas established earlier. The proposed architecture enables online adaptive adjustment of both feedback and coupling gains without requiring any additional communication beyond the states of neighboring agents. Furthermore, we extend our findings to dynamic event-triggered adaptive tracking control, ensuring that Zeno behavior is avoided. Unlike similar adaptive tracking control studies that design feedback or coupling gains exclusively for homogeneous or heterogeneous dynamics, our algorithms account for multiple adaptive gains in heterogeneous and uncertain dynamics, thereby eliminating the need for a distributed observer. Lastly, we provide a numerical example to validate our theoretical algorithms.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"794-804"},"PeriodicalIF":3.0,"publicationDate":"2025-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144750941","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}

{"title":"Multimodal Graph Convolutional Network on Brain Structure and Function in Adolescent Anxiety and Depression","authors":"Sébastien Dam;Jean-Marie Batail;Pierre Maurel;Julie Coloigner","doi":"10.1109/TSIPN.2025.3577354","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3577354","url":null,"abstract":"Multimodal analysis of Magnetic Resonance Imaging (MRI) data enables leveraging complementary information across multiple imaging modalities that may be incomplete when using a single modality. For brain connectivity analysis, graph-based methods, such as graph signal processing, are effective for capturing topological characteristics of the brain structure while incorporating neural activity signals. However, for tasks like group classification, these methods often rely on traditional machine learning algorithms, which may not fully exploit the underlying graph topology. Recently, Graph Convolutional Networks (GCN) have emerged as a powerful tool in brain connectivity research, uncovering complex nonlinear relationships within the data. Here, we develop a multimodal GCN model to jointly model brain structure and function to classify anxiety and depression in adolescents using the Boston Adolescent Neuroimaging of Depression and Anxiety dataset. The graph’s topology is initialized from structural connectivity derived from diffusion MRI, while functional connectivity is incorporated as node features to improve distinction between anxious, depressed patients and healthy controls. Interpretation of key brain regions contributing to classification is enabled through Gradient-weighted Class Activation Mapping, revealing the influence of the frontal and limbic lobes in the diagnosis of the conditions, which aligns with previous findings in the literature. By comparing classification results and the most discriminative features between multimodal and unimodal GCN-based approaches, we demonstrate that our framework improves accuracy in most classification tasks and reveals significant patterns of brain alterations associated with anxiety and depression.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"617-629"},"PeriodicalIF":3.0,"publicationDate":"2025-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144597895","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}

{"title":"Learning Optimal Graph Filters for Clustering of Attributed Graphs","authors":"Meiby Ortiz-Bouza;Selin Aviyente","doi":"10.1109/TSIPN.2025.3574855","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3574855","url":null,"abstract":"Many real-world systems can be represented as graphs where the different entities in the system are presented by nodes and their interactions by edges. An important task in studying large datasets with graphical structure is graph clustering. While there has been a lot of work on graph clustering using the connectivity between the nodes, many real-world networks also have node attributes. Clustering attributed graphs requires joint modeling of graph structure and node attributes. Recent work has focused on combining these two complementary sources of information through graph convolutional networks and graph filtering. However, these methods are mostly limited to lowpass filtering and do not explicitly learn the filter parameters for the clustering task. In this paper, we introduce a graph signal processing based approach, where we learn the parameters of Finite Impulse Response (FIR) and Autoregressive Moving Average (ARMA) graph filters optimized for clustering. The proposed approach is formulated as a two-step iterative optimization problem, focusing on learning interpretable graph filters that are optimal for the given data and that maximize the separation between different clusters. The proposed approach is evaluated on attributed networks and compared to the state-of-the-art methods.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"520-534"},"PeriodicalIF":3.0,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11021005","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144264270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dual-Protection Method Against Eavesdroppers for Distributed State Estimation","authors":"Yan Yu;Chao Yang;Wenjie Ding;Wen Yang;Xiaofan Wang","doi":"10.1109/TSIPN.2025.3563774","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3563774","url":null,"abstract":"This paper examines a security issue for state estimation over a wireless sensor network. The state estimates are transmitted among neighboring nodes through wireless channels in a distributed network, wherein the transmission of the data are vulnerable to the intercept from eavesdroppers, leading to important data privacy leakage. To prevent eavesdroppers from obtaining state estimates, we propose a dual-protection method that combines dynamic transformation with lightweight encryption, which aims to protect the privacy without raising suspicion from eavesdroppers. Furthermore, we consider the scenarios where eavesdroppers utilize side-channel information to gather data and attempt to deduce the encryption mechanism, subsequently inferring the real state estimate. We also provide the analysis to show that the eavesdropper with inference capabilities could not influence the estimation performance of sensors. Finally, the numerical examples are provided to illustrate the effectiveness of the privacy-preserving method.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"427-438"},"PeriodicalIF":3.0,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143913408","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}

{"title":"Resilience in Online Federated Learning: Mitigating Model-Poisoning Attacks via Partial Sharing","authors":"Ehsan Lari;Reza Arablouei;Vinay Chakravarthi Gogineni;Stefan Werner","doi":"10.1109/TSIPN.2025.3559444","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3559444","url":null,"abstract":"Federated learning (FL) allows training machine learning models on distributed data without compromising privacy. However, FL is vulnerable to model-poisoning attacks where malicious clients tamper with their local models to manipulate the global model. In this work, we investigate the resilience of the partial-sharing online FL (PSO-Fed) algorithm against such attacks. PSO-Fed reduces communication overhead by allowing clients to share only a fraction of their model updates with the server. We demonstrate that this partial sharing mechanism has the added advantage of enhancing PSO-Fed's robustness to model-poisoning attacks. Through theoretical analysis, we show that PSO-Fed maintains convergence even under Byzantine attacks, where malicious clients inject noise into their updates. Furthermore, we derive a formula for PSO-Fed's mean square error, considering factors like stepsize, attack probability, and the number of malicious clients. Interestingly, we find a non-trivial optimal stepsize that maximizes PSO-Fed's resistance to these attacks. Extensive numerical experiments confirm our theoretical findings and showcase PSO-Fed's superior performance against model-poisoning attacks compared to other leading FL algorithms.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"388-400"},"PeriodicalIF":3.0,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143883475","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}