Pub Date : 2025-09-24DOI: 10.1109/TSIPN.2025.3613875
Linyao Cao;Wenhua Gao;Jiahong Zhao
Solving distributed optimization problems relies on information exchange between nodes in multi-agent networks. In an unreliable network environment with malicious attacks, compromised nodes deliberately disseminate falsified data to disrupt the optimization process. The security and robustness of the multi-agent system can be improved by designing the fault-tolerant mechanism (FTM) and the resilient distributed optimization (RDO) algorithm. This paper introduces a new fault-tolerant mechanism based on K-Medoids clustering (M-FTM) to address the challenges posed by malicious attacks. Compared with the existing $ F$-local filtering mechanism, M-FTM reduces the network connectivity requirement from $ (2F +1)$-robust to $ (F +1)$-robust, where $ F$ is the number of malicious nodes in the network. This article addresses high-dimensional optimization problems, for which the resilient DIGing algorithm and the resilient Push-DIGing algorithm with fixed step size are proposed. The effectiveness of the algorithms is verified through consensus and convergence analysis. Numerical experiments show that the proposed algorithms can effectively resist malicious attacks. Additionally, M-FTM not only doubles the runtime efficiency of algorithm but also enables its operation under low network connectivity conditions.
{"title":"Resilient Distributed Optimization Algorithm With Fixed Step Size Against Malicious Attacks","authors":"Linyao Cao;Wenhua Gao;Jiahong Zhao","doi":"10.1109/TSIPN.2025.3613875","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3613875","url":null,"abstract":"Solving distributed optimization problems relies on information exchange between nodes in multi-agent networks. In an unreliable network environment with malicious attacks, compromised nodes deliberately disseminate falsified data to disrupt the optimization process. The security and robustness of the multi-agent system can be improved by designing the fault-tolerant mechanism (FTM) and the resilient distributed optimization (RDO) algorithm. This paper introduces a new fault-tolerant mechanism based on K-Medoids clustering (M-FTM) to address the challenges posed by malicious attacks. Compared with the existing <inline-formula><tex-math>$ F$</tex-math></inline-formula>-local filtering mechanism, M-FTM reduces the network connectivity requirement from <inline-formula><tex-math>$ (2F +1)$</tex-math></inline-formula>-robust to <inline-formula><tex-math>$ (F +1)$</tex-math></inline-formula>-robust, where <inline-formula><tex-math>$ F$</tex-math></inline-formula> is the number of malicious nodes in the network. This article addresses high-dimensional optimization problems, for which the resilient DIGing algorithm and the resilient Push-DIGing algorithm with fixed step size are proposed. The effectiveness of the algorithms is verified through consensus and convergence analysis. Numerical experiments show that the proposed algorithms can effectively resist malicious attacks. Additionally, M-FTM not only doubles the runtime efficiency of algorithm but also enables its operation under low network connectivity conditions.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1278-1285"},"PeriodicalIF":3.0,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145210116","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}
Pub Date : 2025-09-24DOI: 10.1109/TSIPN.2025.3613951
Óscar Escudero-Arnanz;Cristina Soguero-Ruiz;Antonio G. Marques
In this paper, we present XST-GCNN (eXplainable Spatio-Temporal Graph Convolutional Neural Network), an innovative architecture designed for processing heterogeneous and irregular Multivariate Time Series (MTS) data. Our processing architecture captures both temporal and feature dependencies within a unified spatio-temporal pipeline by leveraging a GCNN that uses a spatio-temporal graph and aims at optimizing predictive performance and explainability. For graph estimation, we propose several techniques, including a novel approach based on the (heterogeneous) Gower distance. Once the graphs are estimated, we propose two approaches for graph construction: one based on the Cartesian product that treats temporal instants homogeneously, and a spatio-temporal approach that considers different graphs per time step. Finally, we propose two GCNN architectures: a standard GCNN with a normalized adjacency matrix and a higher-order polynomial GCNN. In addition to predictive performance, we incorporate intrinsic explainability through architectural design choices, complemented by post hoc analysis using GNNExplainer, aimed at identifying key feature-time combinations that drive the model’s predictions. We evaluate XST-GCNN using real-world Electronic Health Record data from the University Hospital of Fuenlabrada to predict Multidrug Resistance (MDR) in Intensive Care Unit patients, a critical healthcare challenge associated with high mortality and complex treatments. Our architecture outperforms traditional models, achieving a mean Receiver Operating Characteristic Area Under the Curve score of $mathbf{81.03} pm mathbf{2.43}$. Additionally, the explainability analysis provides actionable insights into clinical factors driving MDR predictions, enhancing model transparency and trust. This work sets a new benchmark for addressing complex inference tasks with heterogeneous and irregular MTS, offering a versatile and interpretable solution for real-world applications.
{"title":"Explainable Spatio-Temporal GCNNs for Irregular Multivariate Time Series: Architecture and Application to ICU Patient Data","authors":"Óscar Escudero-Arnanz;Cristina Soguero-Ruiz;Antonio G. Marques","doi":"10.1109/TSIPN.2025.3613951","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3613951","url":null,"abstract":"In this paper, we present XST-GCNN (eXplainable Spatio-Temporal Graph Convolutional Neural Network), an innovative architecture designed for processing heterogeneous and irregular Multivariate Time Series (MTS) data. Our processing architecture captures both temporal and feature dependencies within a unified spatio-temporal pipeline by leveraging a GCNN that uses a spatio-temporal graph and aims at optimizing predictive performance and explainability. For graph estimation, we propose several techniques, including a novel approach based on the (heterogeneous) Gower distance. Once the graphs are estimated, we propose two approaches for graph construction: one based on the Cartesian product that treats temporal instants homogeneously, and a spatio-temporal approach that considers different graphs per time step. Finally, we propose two GCNN architectures: a standard GCNN with a normalized adjacency matrix and a higher-order polynomial GCNN. In addition to predictive performance, we incorporate intrinsic explainability through architectural design choices, complemented by post hoc analysis using GNNExplainer, aimed at identifying key feature-time combinations that drive the model’s predictions. We evaluate XST-GCNN using real-world Electronic Health Record data from the University Hospital of Fuenlabrada to predict Multidrug Resistance (MDR) in Intensive Care Unit patients, a critical healthcare challenge associated with high mortality and complex treatments. Our architecture outperforms traditional models, achieving a mean Receiver Operating Characteristic Area Under the Curve score of <inline-formula><tex-math>$mathbf{81.03} pm mathbf{2.43}$</tex-math></inline-formula>. Additionally, the explainability analysis provides actionable insights into clinical factors driving MDR predictions, enhancing model transparency and trust. This work sets a new benchmark for addressing complex inference tasks with heterogeneous and irregular MTS, offering a versatile and interpretable solution for real-world applications.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1286-1301"},"PeriodicalIF":3.0,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11178245","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145255936","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}
Pub Date : 2025-09-22DOI: 10.1109/TSIPN.2025.3612840
Xiasheng Shi;Yanxu Su;Darong Huang;Xuyang Lou
This work investigates the economic dispatch problem (EDP) when communication links suffer heterogeneous delays, adopting general convex cost models that subsume the quadratic forms used in existing works. Additionally, each generator is characterized by a bounded output limitation. A distributed primal-dual method is developed through the local communication among generators to address the global equality coupled constraint across all generators. Given the inherent delays in open communication environments, passivity-based scattering transformation is then provided to tackle heterogeneous time delays. Furthermore, we introduce the adaptive penalty function method to handle private bounded constraints. A fully distributed optimization scheme is then developed building on the Karush-Kuhn-Tucker (KKT) conditions to address the EDP over a connected network. We prove its convergence using Filippov solution and differential inclusion. Furthermore, a modified distributed optimization approach incorporating a maximal projection operator is proposed to account for transmission losses. The effectiveness of our developed methods is demonstrated via two practical case simulations.
{"title":"Distributed Optimization Algorithm for the Economic Dispatch Problem Over Delayed Communication Network","authors":"Xiasheng Shi;Yanxu Su;Darong Huang;Xuyang Lou","doi":"10.1109/TSIPN.2025.3612840","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3612840","url":null,"abstract":"This work investigates the economic dispatch problem (EDP) when communication links suffer heterogeneous delays, adopting general convex cost models that subsume the quadratic forms used in existing works. Additionally, each generator is characterized by a bounded output limitation. A distributed primal-dual method is developed through the local communication among generators to address the global equality coupled constraint across all generators. Given the inherent delays in open communication environments, passivity-based scattering transformation is then provided to tackle heterogeneous time delays. Furthermore, we introduce the adaptive penalty function method to handle private bounded constraints. A fully distributed optimization scheme is then developed building on the Karush-Kuhn-Tucker (KKT) conditions to address the EDP over a connected network. We prove its convergence using Filippov solution and differential inclusion. Furthermore, a modified distributed optimization approach incorporating a maximal projection operator is proposed to account for transmission losses. The effectiveness of our developed methods is demonstrated via two practical case simulations.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1332-1343"},"PeriodicalIF":3.0,"publicationDate":"2025-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145351971","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}
Pub Date : 2025-09-17DOI: 10.1109/TSIPN.2025.3611242
Theodor-Adrian Badea;Bogdan Dumitrescu
This paper introduces a novel Laplacian matrix aiming to enable the construction of spectral convolutional networks and to extend the signal processing applications for directed graphs. Our proposal is inspired by a Haar-like transformation and produces a Hermitian matrix which is not only in one-to-one relation with the adjacency matrix, preserving both direction and weight information, but also enjoys desirable additional properties like scaling robustness, sensitivity, continuity, and directionality. We take a theoretical standpoint and support the conformity of our approach with spectral graph theory. Then, we address two use cases: graph learning (by introducing HaarNet, a spectral graph convolutional network built with our Haar-Laplacian) and graph signal processing. We show that our approach gives better results in applications like weight prediction and denoising on directed graphs.
{"title":"Haar-Laplacian for Directed Graphs","authors":"Theodor-Adrian Badea;Bogdan Dumitrescu","doi":"10.1109/TSIPN.2025.3611242","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3611242","url":null,"abstract":"This paper introduces a novel Laplacian matrix aiming to enable the construction of spectral convolutional networks and to extend the signal processing applications for directed graphs. Our proposal is inspired by a Haar-like transformation and produces a Hermitian matrix which is not only in one-to-one relation with the adjacency matrix, preserving both direction and weight information, but also enjoys desirable additional properties like scaling robustness, sensitivity, continuity, and directionality. We take a theoretical standpoint and support the conformity of our approach with spectral graph theory. Then, we address two use cases: graph learning (by introducing HaarNet, a spectral graph convolutional network built with our Haar-Laplacian) and graph signal processing. We show that our approach gives better results in applications like weight prediction and denoising on directed graphs.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1238-1253"},"PeriodicalIF":3.0,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145210115","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}
Pub Date : 2025-09-17DOI: 10.1109/TSIPN.2025.3611264
Victor M. Tenorio;Antonio G. Marques
As the volume and complexity of modern datasets continue to increase, there is an urgent need to develop deep-learning architectures that can process such data efficiently. Graph neural networks (GNNs) have emerged as a promising solution for unstructured data, often outperforming traditional deep-learning models. However, most existing GNNs are designed for a single graph, which limits their applicability in real-world scenarios where multiple graphs may be involved. To address this limitation, we propose a graph-based architecture for tasks in which two sets of signals exist, each defined on a different graph. We first study the supervised and semi-supervised cases, where the input is a signal on one graph (the input graph) and the output is a signal on another graph (the output graph). Our three-block design (i) processes the input graph with a GNN, (ii) applies a latent-space transformation that maps representations from the input to the output graph, and (iii) uses a second GNN that operates on the output graph. Rather than fixing a single implementation for each block, we provide a flexible framework that can be adapted to a variety of problems. The second part of the paper considers a self-supervised setting. Inspired by canonical correlation analysis, we turn our attention to the latent space, seeking informative representations that benefit downstream tasks. By leveraging information from both graphs, the proposed architecture captures richer relationships among entities, leading to improved performance across synthetic and real-world benchmarks. Experiments show consistent gains over conventional deep-learning baselines, highlighting the value of exploiting the two graphs inherent to the task.
{"title":"Exploiting the Structure of Two Graphs With Graph Neural Networks","authors":"Victor M. Tenorio;Antonio G. Marques","doi":"10.1109/TSIPN.2025.3611264","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3611264","url":null,"abstract":"As the volume and complexity of modern datasets continue to increase, there is an urgent need to develop deep-learning architectures that can process such data efficiently. Graph neural networks (GNNs) have emerged as a promising solution for unstructured data, often outperforming traditional deep-learning models. However, most existing GNNs are designed for a single graph, which limits their applicability in real-world scenarios where multiple graphs may be involved. To address this limitation, we propose a graph-based architecture for tasks in which two sets of signals exist, each defined on a different graph. We first study the supervised and semi-supervised cases, where the input is a signal on one graph (the <italic>input graph</i>) and the output is a signal on another graph (the <italic>output graph</i>). Our three-block design (i) processes the input graph with a GNN, (ii) applies a latent-space transformation that maps representations from the input to the output graph, and (iii) uses a second GNN that operates on the output graph. Rather than fixing a single implementation for each block, we provide a flexible framework that can be adapted to a variety of problems. The second part of the paper considers a self-supervised setting. Inspired by canonical correlation analysis, we turn our attention to the latent space, seeking informative representations that benefit downstream tasks. By leveraging information from both graphs, the proposed architecture captures richer relationships among entities, leading to improved performance across synthetic and real-world benchmarks. Experiments show consistent gains over conventional deep-learning baselines, highlighting the value of exploiting the two graphs inherent to the task.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1254-1267"},"PeriodicalIF":3.0,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11168903","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145255939","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}
Pub Date : 2025-09-17DOI: 10.1109/TSIPN.2025.3611172
Eugenio Borzone;Leandro Di Persia;Matias Gerard
This paper presents a novelgraph-based deep learning model for tasks involving relations between two nodes (edge-centric tasks), where the focuslies on predicting relationships and interactions between pairs of nodes rather than node properties themselves. This model combines supervised and self-supervisedlearning, taking into account for the loss function the embeddings learned and patterns with and without ground truth. Additionally it incorporates an attention mechanism that leveragesboth node and edge features. The architecture, trained end-to-end, comprises two primary components: embedding generation and prediction. First, a graph neural network (GNN) transform raw node features into dense, low-dimensional embeddings, incorporating edge attributes. Then, a feedforward neural model processes the node embeddings to produce the final output. Experiments demonstrate that our model matches or exceeds existing methods for protein-protein interactions prediction and Gene Ontology (GO) terms prediction. The model also performs effectively with one-hot encoding for node features, providing a solution for the previously unsolved problem of predicting similarity between compounds with unknown structures.
{"title":"A Hybrid Supervised and Self-Supervised Graph Neural Network for Edge-Centric Applications","authors":"Eugenio Borzone;Leandro Di Persia;Matias Gerard","doi":"10.1109/TSIPN.2025.3611172","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3611172","url":null,"abstract":"This paper presents a novelgraph-based deep learning model for tasks involving relations between two nodes (edge-centric tasks), where the focuslies on predicting relationships and interactions between pairs of nodes rather than node properties themselves. This model combines supervised and self-supervisedlearning, taking into account for the loss function the embeddings learned and patterns with and without ground truth. Additionally it incorporates an attention mechanism that leveragesboth node and edge features. The architecture, trained end-to-end, comprises two primary components: embedding generation and prediction. First, a graph neural network (GNN) transform raw node features into dense, low-dimensional embeddings, incorporating edge attributes. Then, a feedforward neural model processes the node embeddings to produce the final output. Experiments demonstrate that our model matches or exceeds existing methods for protein-protein interactions prediction and Gene Ontology (GO) terms prediction. The model also performs effectively with one-hot encoding for node features, providing a solution for the previously unsolved problem of predicting similarity between compounds with unknown structures.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1268-1277"},"PeriodicalIF":3.0,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145210114","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}
Pub Date : 2025-09-11DOI: 10.1109/TSIPN.2025.3608959
Chang Ye;Gonzalo Mateos
Starting from first graph signal processing (GSP) principles, we present a novel model-based deep learning approach to blind deconvolution of sparse graph signals. Despite the bilinear nature of the observations, by requiring invertibility of the unknown (diffusion graph filter) forward operator we can formulate a convex optimization problem and solve it using the alternating-direction method of multipliers (ADMM). We then unroll and truncate the novel ADMM iterations to arrive at a parameterized neural network architecture for blind deconvolution on graphs (BDoG-Net), which we train in an end-to-end fashion using labeled data. This supervised learning approach offers several advantages, such as interpretability, parameter efficiency, and controllable complexity during inference. Our reproducible numerical experiments corroborate that BDoG-Net exhibits performance on par with the iterative ADMM baseline, but with markedly faster inference times and without the need to manually adjust the step-size or penalty parameters. The application of BDoG-Net to a simplified instance of source localization over networks is also discussed. Overall, our approach combines the best of both worlds by incorporating the inductive biases of a GSP model-based solution within a data-driven, trainable deep learning architecture for blind deconvolution on graphs.
{"title":"BDoG-Net: Algorithm Unrolling for Blind Deconvolution on Graphs","authors":"Chang Ye;Gonzalo Mateos","doi":"10.1109/TSIPN.2025.3608959","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3608959","url":null,"abstract":"Starting from first graph signal processing (GSP) principles, we present a novel model-based deep learning approach to blind deconvolution of sparse graph signals. Despite the bilinear nature of the observations, by requiring invertibility of the unknown (diffusion graph filter) forward operator we can formulate a convex optimization problem and solve it using the alternating-direction method of multipliers (ADMM). We then unroll and truncate the novel ADMM iterations to arrive at a parameterized neural network architecture for blind deconvolution on graphs (BDoG-Net), which we train in an end-to-end fashion using labeled data. This supervised learning approach offers several advantages, such as interpretability, parameter efficiency, and controllable complexity during inference. Our reproducible numerical experiments corroborate that BDoG-Net exhibits performance on par with the iterative ADMM baseline, but with markedly faster inference times and without the need to manually adjust the step-size or penalty parameters. The application of BDoG-Net to a simplified instance of source localization over networks is also discussed. Overall, our approach combines the best of both worlds by incorporating the inductive biases of a GSP model-based solution within a data-driven, trainable deep learning architecture for blind deconvolution on graphs.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1200-1213"},"PeriodicalIF":3.0,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145110327","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}
Pub Date : 2025-09-11DOI: 10.1109/TSIPN.2025.3608945
Huan Li;Shuangsi Xue;Zihang Guo;Junkai Tan;Hui Cao;Dongyu Li
This article investigates a special multi-agent group consensus control problem—the restricted-sequence-synchronized (RSS) group consensus, where all subgroups achieve their respective consensus according to a restricted group consensus sequence, while agents within each subgroup simultaneously reach consensus. To comprehensively express this problem, we first introduce RSS stability, where for a single system, all of its state components arrive at the stable state following a restricted sequence. Next, the concept of RSS stability, initially applied to a single system, is extended to the RSS group consensus of multi-agent systems. Furthermore, a sliding-mode control protocol is devised to achieve RSS group consensus in heterogeneous multi-agent systems and handle the practical impact of actuator faults and external disturbance. Adaptive techniques are incorporated within this RSS group consensus controller to dynamically address the actuator faults. Two simulation cases illustrate the effective performance of the developed RSS group consensus control protocol.
{"title":"Sequence-Based Group Consensus for Heterogeneous Multi-Agent Systems","authors":"Huan Li;Shuangsi Xue;Zihang Guo;Junkai Tan;Hui Cao;Dongyu Li","doi":"10.1109/TSIPN.2025.3608945","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3608945","url":null,"abstract":"This article investigates a special multi-agent group consensus control problem—the restricted-sequence-synchronized (RSS) group consensus, where all subgroups achieve their respective consensus according to a restricted group consensus sequence, while agents within each subgroup simultaneously reach consensus. To comprehensively express this problem, we first introduce RSS stability, where for a single system, all of its state components arrive at the stable state following a restricted sequence. Next, the concept of RSS stability, initially applied to a single system, is extended to the RSS group consensus of multi-agent systems. Furthermore, a sliding-mode control protocol is devised to achieve RSS group consensus in heterogeneous multi-agent systems and handle the practical impact of actuator faults and external disturbance. Adaptive techniques are incorporated within this RSS group consensus controller to dynamically address the actuator faults. Two simulation cases illustrate the effective performance of the developed RSS group consensus control protocol.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1188-1199"},"PeriodicalIF":3.0,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145110293","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}
In EEG-based connectivity analysis, multiple graphs are typically available, as obtained from measurements taken under different conditions, such as frequency bands, trials, or patients’ states. Identifying stable patterns of interaction among brain regions, such as functional communities, can provide valuable insight into brain organization and its changes across tasks or conditions. It is then of interest to find a method to properly combine these different graphs to obtain a clustering that is sufficiently stable and representative of brain functionalities. In this paper, we propose a method to obtain robust spectral clustering. The method relies on a statistical characterization of the multiple graphs and a small perturbation analysis of the eigen-decomposition of graphs affected by random perturbations to derive the optimal weighting that minimizes the variance of the eigenvalues. The proposed method is first tested on synthetic data, to assess its advantages with respect to conventional approaches under controllable conditions, and then it is applied to real EEG data in both healthy individuals performing motor imagery tasks and patients affected by Alzheimer’s disease. The results show the ability of the method to reliably detect functional communities and quantify connectivity reorganization during cognitive tasks. Our results suggest that the proposed approach provides a valid new strategy to combine multiple graphs taking into account the statistical properties of each graph in the presence of uncertainties.
{"title":"Robust Graph Topology Inference for Multiple Brain EEG Networks","authors":"Tiziana Cattai;Stefania Colonnese;Sergio Barbarossa","doi":"10.1109/TSIPN.2025.3606164","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3606164","url":null,"abstract":"In EEG-based connectivity analysis, multiple graphs are typically available, as obtained from measurements taken under different conditions, such as frequency bands, trials, or patients’ states. Identifying stable patterns of interaction among brain regions, such as functional communities, can provide valuable insight into brain organization and its changes across tasks or conditions. It is then of interest to find a method to properly combine these different graphs to obtain a clustering that is sufficiently stable and representative of brain functionalities. In this paper, we propose a method to obtain robust spectral clustering. The method relies on a statistical characterization of the multiple graphs and a small perturbation analysis of the eigen-decomposition of graphs affected by random perturbations to derive the optimal weighting that minimizes the variance of the eigenvalues. The proposed method is first tested on synthetic data, to assess its advantages with respect to conventional approaches under controllable conditions, and then it is applied to real EEG data in both healthy individuals performing motor imagery tasks and patients affected by Alzheimer’s disease. The results show the ability of the method to reliably detect functional communities and quantify connectivity reorganization during cognitive tasks. Our results suggest that the proposed approach provides a valid new strategy to combine multiple graphs taking into account the statistical properties of each graph in the presence of uncertainties.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1317-1331"},"PeriodicalIF":3.0,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11157831","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145352019","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}
Pub Date : 2025-09-04DOI: 10.1109/TSIPN.2025.3606196
Yuwei Ren;Putian Cai;Yixian Fang;Ben Niu
This article investigates a dynamic quantized event-triggered predictive control policy to stabilize a linear system with denial-of-service attacks. First, to address the challenges of quantization errors and DoS attacks, a co-design approach integrating event-triggered control and predictive control is proposed to ensure the stability of networked control systems. Second, a novel model framework is developed, which combines a dynamic quantizer with asynchronous event-triggered control mechanisms for practical implementation. Subsequently, a new hybrid system framework is adopted for modeling closed-loop dynamics. Using Lyapunov theory, the input-to-state stability of the closed-loop system is guaranteed through derived sufficient conditions with constrains of quantization parameters and event-triggered mechanisms. Finally, the presented example validates the effectiveness of the transmission policy proposed in this article.
{"title":"Dynamic Quantized Event-Triggered Predictive Control for Networked Control Systems With DoS Attacks: A Hybrid System Approach","authors":"Yuwei Ren;Putian Cai;Yixian Fang;Ben Niu","doi":"10.1109/TSIPN.2025.3606196","DOIUrl":"https://doi.org/10.1109/TSIPN.2025.3606196","url":null,"abstract":"This article investigates a dynamic quantized event-triggered predictive control policy to stabilize a linear system with denial-of-service attacks. First, to address the challenges of quantization errors and DoS attacks, a co-design approach integrating event-triggered control and predictive control is proposed to ensure the stability of networked control systems. Second, a novel model framework is developed, which combines a dynamic quantizer with asynchronous event-triggered control mechanisms for practical implementation. Subsequently, a new hybrid system framework is adopted for modeling closed-loop dynamics. Using Lyapunov theory, the input-to-state stability of the closed-loop system is guaranteed through derived sufficient conditions with constrains of quantization parameters and event-triggered mechanisms. Finally, the presented example validates the effectiveness of the transmission policy proposed in this article.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"11 ","pages":"1138-1150"},"PeriodicalIF":3.0,"publicationDate":"2025-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145051049","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：