Pub Date : 2025-01-07DOI: 10.1109/TNSE.2024.3524616
Lulu Li;Huihui Zhang;Daniel W. C. Ho
This paper investigates the consensus issue of multi-agent systems under constrained data rates, time delays, and denial-of-service (DoS) attacks. We first introduce a periodically adjusted dynamic quantizer based on the equally distributed bit rate model, which can effectively avoid saturation and eliminate the quantization error over time, unlike the static quantizer. Then, we show that the quantizer in this paper is suitable for multi-agent systems with time delays, and we design a quantized controller that can realize the consensus in such systems. We also derive the sufficient bit rate condition for achieving consensus under time delays. Next, we extend our approach to handle multi-agent systems with both time delays and DoS attacks under the general energy-constrained DoS model. We provide the conditions on bit rate and average duration and frequency of DoS attacks that ensure system performance. Finally, we analyze the relationship between system performance, bit rate, time delays, and DoS attacks, and verify our results by numerical examples.
{"title":"Delay and DoS Resilient Consensus of Multi-Agent Systems: A Bit Rate Minimization Strategy","authors":"Lulu Li;Huihui Zhang;Daniel W. C. Ho","doi":"10.1109/TNSE.2024.3524616","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3524616","url":null,"abstract":"This paper investigates the consensus issue of multi-agent systems under constrained data rates, time delays, and denial-of-service (DoS) attacks. We first introduce a periodically adjusted dynamic quantizer based on the equally distributed bit rate model, which can effectively avoid saturation and eliminate the quantization error over time, unlike the static quantizer. Then, we show that the quantizer in this paper is suitable for multi-agent systems with time delays, and we design a quantized controller that can realize the consensus in such systems. We also derive the sufficient bit rate condition for achieving consensus under time delays. Next, we extend our approach to handle multi-agent systems with both time delays and DoS attacks under the general energy-constrained DoS model. We provide the conditions on bit rate and average duration and frequency of DoS attacks that ensure system performance. Finally, we analyze the relationship between system performance, bit rate, time delays, and DoS attacks, and verify our results by numerical examples.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"1159-1171"},"PeriodicalIF":6.7,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465862","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}
Pub Date : 2025-01-06DOI: 10.1109/TNSE.2024.3523952
Zhigang Wang;Ye Deng;Ze Wang;Jürgen Kurths;Jun Wu
Many real complex systems, such as infrastructure and the Internet, are not random but embedded in a metric space. The problem of spatial network disintegration, or critical area identification, is a fundamental research domain in network science and has received increasing attention. Typical applications include network immunization, epidemic control, and early warning signals of disintegration. Due to the computationally challenging (NP-hard) problem, they usually cannot be solved with polynomial algorithms. Here, we propose an efficient disintegration method in spatial networks through a link-based strategy. First, we introduce a regional failure model with multiple disintegration circles for the spatial network. We then calculate the sum of the specific attribute values of the links in the circle to identify the critical regions of the spatial network, which also correspond to the geographic regions where disintegration occurs. Extensive experiments on real-world networks of different types demonstrate that the strategy outperforms conventional methods in terms of solution quality.
{"title":"Efficient Link-Based Spatial Network Disintegration Strategy","authors":"Zhigang Wang;Ye Deng;Ze Wang;Jürgen Kurths;Jun Wu","doi":"10.1109/TNSE.2024.3523952","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3523952","url":null,"abstract":"Many real complex systems, such as infrastructure and the Internet, are not random but embedded in a metric space. The problem of spatial network disintegration, or critical area identification, is a fundamental research domain in network science and has received increasing attention. Typical applications include network immunization, epidemic control, and early warning signals of disintegration. Due to the computationally challenging (NP-hard) problem, they usually cannot be solved with polynomial algorithms. Here, we propose an efficient disintegration method in spatial networks through a link-based strategy. First, we introduce a regional failure model with multiple disintegration circles for the spatial network. We then calculate the sum of the specific attribute values of the links in the circle to identify the critical regions of the spatial network, which also correspond to the geographic regions where disintegration occurs. Extensive experiments on real-world networks of different types demonstrate that the strategy outperforms conventional methods in terms of solution quality.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"1096-1111"},"PeriodicalIF":6.7,"publicationDate":"2025-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465721","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}
In online social networks, numerous studies have demonstrated the challenge of predicting who will eventually engage in an information cascade with its initial part. Take a step back. Can we predict who will engage in the cascade at the next stage if the lifetime of cascades is divided into a certain number of stages? Although numerous attempts have been made to solve this problem, how to extract useful information from the historical cascades spreading within a sub-network and the connections among users remains an open question. This paper proposes a simple but efficient unsupervised agent-based model, the triple ranking model, which integrates exposure time ranking, social gravity ranking, and cascade similarity ranking. The rankings, a key component of our model, have been successful in characterizing the social impact of shifted users, temporal information, and sequential cascade information, demonstrating the generalizability of our approach. To test the contributions of the features in supervised frameworks, we fuse them with two graph neural networks, the graph convolutional network (GCN) and graph attention network (GAT). Our experimental results on three Twitter networks unequivocally show that the proposed algorithm outperforms the tested state-of-art algorithms across a series of performance metrics. Notably, its time complexity is also lower than theirs, further underscoring its superiority. The observations demonstrate that the rankings effectively abstract the features hidden in the information cascades and in the topology of social networks, paving the way for further studies on posting engagement.
{"title":"Predicting Participation Shift of Users at the Next Stage in Social Networks","authors":"Yichao Zhang;Zejian Wang;Huangxin Zhuang;Lei Song;Guanghui Wen;Jihong Guan;Shuigeng Zhou","doi":"10.1109/TNSE.2024.3523300","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3523300","url":null,"abstract":"In online social networks, numerous studies have demonstrated the challenge of predicting who will eventually engage in an information cascade with its initial part. Take a step back. Can we predict who will engage in the cascade at the next stage if the lifetime of cascades is divided into a certain number of stages? Although numerous attempts have been made to solve this problem, how to extract useful information from the historical cascades spreading within a sub-network and the connections among users remains an open question. This paper proposes a simple but efficient unsupervised agent-based model, the triple ranking model, which integrates exposure time ranking, social gravity ranking, and cascade similarity ranking. The rankings, a key component of our model, have been successful in characterizing the social impact of shifted users, temporal information, and sequential cascade information, demonstrating the generalizability of our approach. To test the contributions of the features in supervised frameworks, we fuse them with two graph neural networks, the graph convolutional network (GCN) and graph attention network (GAT). Our experimental results on three Twitter networks unequivocally show that the proposed algorithm outperforms the tested state-of-art algorithms across a series of performance metrics. Notably, its time complexity is also lower than theirs, further underscoring its superiority. The observations demonstrate that the rankings effectively abstract the features hidden in the information cascades and in the topology of social networks, paving the way for further studies on posting engagement.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"1066-1079"},"PeriodicalIF":6.7,"publicationDate":"2025-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465757","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}
Predicting the popularity of information in social networks poses a highly challenging problem. The popularity of a message is contingent upon its diffusion process and the relationships it maintains with other cascades and is influenced by user behaviour within the social network. Effectively capturing the dynamic process of message propagation and integrating the structural features of the social network to enhance popularity prediction constitutes a pivotal challenge. To address the challenge, we propose a novel method called “Cascade Social Net” (CSN) that leverages cascade graphs and social graphs to predict cascade popularity accurately. The proposed method consists of three stages. Firstly, we construct a social graph by collecting user information and their connections. Secondly, we integrate information from social graphs, cascade graphs and inter-cascade graphs. Finally, we leverage graph neural networks to predict the popularity of cascades. To overcome the challenge of large-scale social graphs, we introduce a novel neighbour sampling technique that efficiently aggregates information from second-order neighbours. We evaluate our method on real-world datasets and compare it with state-of-the-art methods. Our results demonstrate that CSN outperforms existing methods in predicting cascade popularity.
{"title":"Cascade Popularity Prediction: A Multi-View Learning Approach With Socialized Modeling","authors":"Pengfei Jiao;Weijian Song;Yuling Wang;Wang Zhang;Hongqian Chen;Zhidong Zhao;Jian Wu","doi":"10.1109/TNSE.2025.3525717","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3525717","url":null,"abstract":"Predicting the popularity of information in social networks poses a highly challenging problem. The popularity of a message is contingent upon its diffusion process and the relationships it maintains with other cascades and is influenced by user behaviour within the social network. Effectively capturing the dynamic process of message propagation and integrating the structural features of the social network to enhance popularity prediction constitutes a pivotal challenge. To address the challenge, we propose a novel method called “Cascade Social Net” (CSN) that leverages cascade graphs and social graphs to predict cascade popularity accurately. The proposed method consists of three stages. Firstly, we construct a social graph by collecting user information and their connections. Secondly, we integrate information from social graphs, cascade graphs and inter-cascade graphs. Finally, we leverage graph neural networks to predict the popularity of cascades. To overcome the challenge of large-scale social graphs, we introduce a novel neighbour sampling technique that efficiently aggregates information from second-order neighbours. We evaluate our method on real-world datasets and compare it with state-of-the-art methods. Our results demonstrate that CSN outperforms existing methods in predicting cascade popularity.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"1198-1209"},"PeriodicalIF":6.7,"publicationDate":"2025-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465712","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}
Pub Date : 2025-01-01DOI: 10.1109/TNSE.2024.3524576
Han Wang;Anfeng Liu;Neal N. Xiong
Many game-based data collection schemes aim to optimize system profits in Mobile Crowd Sensing (MCS). These schemes often assume that the platform knows the data quality upon receipt from workers, ignoring the verification costs; and assume all low-quality submissions are detected. However, due to the challenge of Information Elicitation Without Verification (IEWV), previous game strategies fail to address two key issues in real-world MCS: (1) Verification incurs costs, so the Nash equilibrium from previous studies may not hold. (2) Cheating workers may not be detected, leading to poor-quality data submissions, a scenario not considered in previous models. To address these challenges, we propose a new Stackelberg Game-based quality Control System (SGCS). Theoretically, we derive the minimum verification rate required for workers to submit high-quality data, considering their strategic responses to the platform's verification rate. We also design a Worker-Dependent Verification Rates (WDVR) algorithm that identifies honest workers focusing on long-term gains, reducing verification rates for them to lower average verification costs and enhance platform utilities. Our approach is validated through a drone-assisted data collection application, demonstrating that: (1) A minimum effective verification rate ensures strategic workers submit high-quality data. (2) There is a complex trade-off between data quality, verification rates, and platform utilities. Higher data quality increases platform income but also raises verification costs more rapidly, potentially reducing overall utilities. The proposed SGCS provides a practical game-theoretic method for MCS data collection.
{"title":"SGCS: A Cost-Effective Quality Control System for Strategic Workers in Mobile Crowd Sensing","authors":"Han Wang;Anfeng Liu;Neal N. Xiong","doi":"10.1109/TNSE.2024.3524576","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3524576","url":null,"abstract":"Many game-based data collection schemes aim to optimize system profits in Mobile Crowd Sensing (MCS). These schemes often assume that the platform knows the data quality upon receipt from workers, ignoring the verification costs; and assume all low-quality submissions are detected. However, due to the challenge of Information Elicitation Without Verification (IEWV), previous game strategies fail to address two key issues in real-world MCS: (1) Verification incurs costs, so the Nash equilibrium from previous studies may not hold. (2) Cheating workers may not be detected, leading to poor-quality data submissions, a scenario not considered in previous models. To address these challenges, we propose a new <underline>S</u>tackelberg <underline>G</u>ame-based quality <underline>C</u>ontrol <underline>S</u>ystem (SGCS). Theoretically, we derive the minimum verification rate required for workers to submit high-quality data, considering their strategic responses to the platform's verification rate. We also design a Worker-Dependent Verification Rates (WDVR) algorithm that identifies honest workers focusing on long-term gains, reducing verification rates for them to lower average verification costs and enhance platform utilities. Our approach is validated through a drone-assisted data collection application, demonstrating that: (1) A minimum effective verification rate ensures strategic workers submit high-quality data. (2) There is a complex trade-off between data quality, verification rates, and platform utilities. Higher data quality increases platform income but also raises verification costs more rapidly, potentially reducing overall utilities. The proposed SGCS provides a practical game-theoretic method for MCS data collection.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"1146-1158"},"PeriodicalIF":6.7,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465580","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}
Pub Date : 2024-12-31DOI: 10.1109/TNSE.2024.3522891
Le Cheng;Peican Zhu;Chao Gao;Zhen Wang;Xuelong Li
The dissemination of rumors imposes substantial hazards. Therefore, accurately identifying the source of rumor and promptly controlling information propagation hold paramount practical significance. Presently, prevailing sensor deployment methods solely focus on network structural information, disregarding the propagation process, thus incurring certain limitations. Additionally, source detection methods presuppose reliable assumptions, i.e., complete network structure and observational data. However, due to temporal constraints and cost considerations, the acquired network information is often structurally incomplete: partial edges missing. To address these issues, this paper introduces a novel approach, namely Source Detection in Structurally Incomplete social networks (SDSI). Firstly, to monitor the network efficiently, a certain number of sensors are deployed using quality-guaranteed Monte Carlo simulations to achieve maximum coverage. In the source detection phase, considering the acquired incomplete information, the source node is determined based on Bayesian posterior maximum estimation. Additionally, SDSI is enhanced through incorporating the sharing counts of the information in social networks. Extensive experiments in diverse scenarios demonstrate the superiority and robustness of the proposed SDSI.
{"title":"SDSI: Source Detection in Structurally Incomplete Social Networks","authors":"Le Cheng;Peican Zhu;Chao Gao;Zhen Wang;Xuelong Li","doi":"10.1109/TNSE.2024.3522891","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3522891","url":null,"abstract":"The dissemination of rumors imposes substantial hazards. Therefore, accurately identifying the source of rumor and promptly controlling information propagation hold paramount practical significance. Presently, prevailing sensor deployment methods solely focus on network structural information, disregarding the propagation process, thus incurring certain limitations. Additionally, source detection methods presuppose reliable assumptions, i.e., complete network structure and observational data. However, due to temporal constraints and cost considerations, the acquired network information is often structurally incomplete: partial edges missing. To address these issues, this paper introduces a novel approach, namely <bold>S</b>ource <bold>D</b>etection in <bold>S</b>tructurally <bold>I</b>ncomplete social networks (SDSI). Firstly, to monitor the network efficiently, a certain number of sensors are deployed using quality-guaranteed Monte Carlo simulations to achieve maximum coverage. In the source detection phase, considering the acquired incomplete information, the source node is determined based on Bayesian posterior maximum estimation. Additionally, SDSI is enhanced through incorporating the sharing counts of the information in social networks. Extensive experiments in diverse scenarios demonstrate the superiority and robustness of the proposed SDSI.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"1041-1052"},"PeriodicalIF":6.7,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465860","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}
Learning from the lessons of the COVID-19 pandemic, nations are increasingly recognizing the imperative to develop sustainable mobility interventions that effectively balance epidemic control and economic stability. In response, we study a novel network immunity problem: the formulation of precise capacity limitation measures for each point of interest (POI) node within the urban mobility network. The aim is to maximize epidemic containment under the fixed resource budget for mobility intervention. To achieve this, we establish a metapopulation model on urban inter-POI networks. Our proposed model accurately fits real epidemic trajectories, demonstrating resilience to significant shifts in human movement patterns pre- and post-epidemic. Leveraging this model, we derive the generalized basic reproduction number and reframe the original problem as one that minimizes $R_{0}$ under budgetary constraints. We devise a greedy capacity reduction algorithm to approximately solve these problems. Subsequently, we conduct extensive experiments on large-scale urban networks that connect 4,335 residential communities to 14,936 POIs with 5.7 million daily edges. Compared to baseline methods, our algorithm consistently achieves higher efficiency and accuracy in reducing $R_{0}$ and maximizing epidemic containment. Notably, it can effectively minimize the risk of epidemic spread within the city without imposing significant constraints on overall urban mobility.
{"title":"Optimizing Sustainable Mobility Interventions for Efficient Epidemic Containment","authors":"Yanggang Cheng;Shibo He;Cunqi Shao;Chao Li;Jiming Chen","doi":"10.1109/TNSE.2024.3519670","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3519670","url":null,"abstract":"Learning from the lessons of the COVID-19 pandemic, nations are increasingly recognizing the imperative to develop sustainable mobility interventions that effectively balance epidemic control and economic stability. In response, we study a novel network immunity problem: the formulation of precise capacity limitation measures for each point of interest (POI) node within the urban mobility network. The aim is to maximize epidemic containment under the fixed resource budget for mobility intervention. To achieve this, we establish a metapopulation model on urban inter-POI networks. Our proposed model accurately fits real epidemic trajectories, demonstrating resilience to significant shifts in human movement patterns pre- and post-epidemic. Leveraging this model, we derive the generalized basic reproduction number and reframe the original problem as one that minimizes <inline-formula><tex-math>$R_{0}$</tex-math></inline-formula> under budgetary constraints. We devise a greedy capacity reduction algorithm to approximately solve these problems. Subsequently, we conduct extensive experiments on large-scale urban networks that connect 4,335 residential communities to 14,936 POIs with 5.7 million daily edges. Compared to baseline methods, our algorithm consistently achieves higher efficiency and accuracy in reducing <inline-formula><tex-math>$R_{0}$</tex-math></inline-formula> and maximizing epidemic containment. Notably, it can effectively minimize the risk of epidemic spread within the city without imposing significant constraints on overall urban mobility.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"823-837"},"PeriodicalIF":6.7,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465772","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}
Pub Date : 2024-12-30DOI: 10.1109/TNSE.2024.3524077
Yue Yang;Guodong Li;Dongxu Li;Jun Zhang;Pengwei Hu;Lun Hu
Attributed graph clustering is of significance for an in-depth understanding of the intrinsic organization of complex networks. Recently, owing to the powerful learning capability of deep neural networks, increasing attention has been paid to developing more advanced deep learning approaches for attributed graph clustering. However, these approaches mainly focus on the purpose of obtaining general node representations, followed by clustering as a downstream task in combination with traditional techniques. Intuitively, the clustering performance is further limited due to the lack of achieving the desirable clustering quality in the graph representation learning process. To this end, this paper proposes a novel end-to-end attributed graph clustering model, namely FCGCN, by integrating fuzzy clustering and graph convolution network. FCGCN is trained toward optimizing an unsupervised fuzzy-based clustering objective, which is specifically formulated for precisely evaluating the clustering quality by considering the fuzzy memberships of nodes over clusters. To avoid the generation of undesirable clusters, we introduce a tailored regularization term by alleviating the over-smoothing issue on graph neural networks. By doing so, an explicit connection between graph representation learning and clustering can thus be established, considerably improving the clustering performance. To evaluate the performance of FCGCN, extensive experiments are conducted on a total of five real attributed graph datasets, and the results demonstrate the superiority of FCGCN over several state-of-the-art algorithms in terms of effectiveness and efficiency.
{"title":"Integrating Fuzzy Clustering and Graph Convolution Network to Accurately Identify Clusters From Attributed Graph","authors":"Yue Yang;Guodong Li;Dongxu Li;Jun Zhang;Pengwei Hu;Lun Hu","doi":"10.1109/TNSE.2024.3524077","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3524077","url":null,"abstract":"Attributed graph clustering is of significance for an in-depth understanding of the intrinsic organization of complex networks. Recently, owing to the powerful learning capability of deep neural networks, increasing attention has been paid to developing more advanced deep learning approaches for attributed graph clustering. However, these approaches mainly focus on the purpose of obtaining general node representations, followed by clustering as a downstream task in combination with traditional techniques. Intuitively, the clustering performance is further limited due to the lack of achieving the desirable clustering quality in the graph representation learning process. To this end, this paper proposes a novel end-to-end attributed graph clustering model, namely FCGCN, by integrating fuzzy clustering and graph convolution network. FCGCN is trained toward optimizing an unsupervised fuzzy-based clustering objective, which is specifically formulated for precisely evaluating the clustering quality by considering the fuzzy memberships of nodes over clusters. To avoid the generation of undesirable clusters, we introduce a tailored regularization term by alleviating the over-smoothing issue on graph neural networks. By doing so, an explicit connection between graph representation learning and clustering can thus be established, considerably improving the clustering performance. To evaluate the performance of FCGCN, extensive experiments are conducted on a total of five real attributed graph datasets, and the results demonstrate the superiority of FCGCN over several state-of-the-art algorithms in terms of effectiveness and efficiency.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"1112-1125"},"PeriodicalIF":6.7,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465775","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}
Pub Date : 2024-12-30DOI: 10.1109/TNSE.2024.3524383
Huarong Zhao;Li Peng;Linbo Xie;Hongnian Yu
This paper addresses the challenge of achieving fully distributed event-triggered bipartite consensus in discrete-time nonlinear multi-agent systems (MASs) characterized by unknown dynamic models and antagonistic interactions. It begins by transforming the bipartite consensus issue into a standard consensus problem through a combined measurement error function. A dynamic linearization model is subsequently established for the input and the combined measurement error function, easing the strongly connected requirement of MASs' communication topology. To enhance performance, an event-triggered model-free sliding-mode bipartite consensus algorithm is proposed, designed to boost convergence speed, reduce steady-state error, and relieve communication burden. The convergence of the proposed method is rigorously proven, allowing for control of fine-tuned specific practical needs. Simulation studies are conducted to verify the effectiveness of the proposed scheme.
{"title":"Event-Triggered Bipartite Consensus for Multi-Agent Systems via Model-Free Sliding-Mode Scheme","authors":"Huarong Zhao;Li Peng;Linbo Xie;Hongnian Yu","doi":"10.1109/TNSE.2024.3524383","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3524383","url":null,"abstract":"This paper addresses the challenge of achieving fully distributed event-triggered bipartite consensus in discrete-time nonlinear multi-agent systems (MASs) characterized by unknown dynamic models and antagonistic interactions. It begins by transforming the bipartite consensus issue into a standard consensus problem through a combined measurement error function. A dynamic linearization model is subsequently established for the input and the combined measurement error function, easing the strongly connected requirement of MASs' communication topology. To enhance performance, an event-triggered model-free sliding-mode bipartite consensus algorithm is proposed, designed to boost convergence speed, reduce steady-state error, and relieve communication burden. The convergence of the proposed method is rigorously proven, allowing for control of fine-tuned specific practical needs. Simulation studies are conducted to verify the effectiveness of the proposed scheme.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"1137-1145"},"PeriodicalIF":6.7,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465698","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}
Pub Date : 2024-12-30DOI: 10.1109/TNSE.2024.3521929
Lin Chen;Congyi Wang;Zhaoyuan Wu
Amidst the rapid transformation of the electricity supply and demand structure, there has been a consensus among policy makers on the need for further refinement of the time-of-use (ToU) pricing mechanism. Nonetheless, the challenge of capturing the dynamic interplay between ToU pricing design and the market behavior of diverse consumers, particularly in light of distributed energy integration, persists as an unresolved inquiry. This paper introduces a novel approach, utilizing reinforcement learning for the development of a ToU pricing model considering investment of distributed energy. The interaction between end consumers and utilities within the ToU pricing framework is encapsulated within a bi-level structure, characterized by significant applicability and scalability. A deep reinforcement learning algorithm is employed to train an agent in devising effective pricing strategies. To aid the agent in grasping the complex, interdependent pricing effects during peak, mid-peak, and valley periods, a configuration employing three interconnected Long Short-Term Memory networks is adopted. Case studies, grounded in empirical datasets, substantiate the efficacy and rationality of the methodology presented herein. It is anticipated that the framework proposed in this study will serve as a valuable reference for the design of efficient ToU pricing in diverse regions.
{"title":"Reinforcement Learning-Based Time of Use Pricing Design Toward Distributed Energy Integration in Low Carbon Power System","authors":"Lin Chen;Congyi Wang;Zhaoyuan Wu","doi":"10.1109/TNSE.2024.3521929","DOIUrl":"https://doi.org/10.1109/TNSE.2024.3521929","url":null,"abstract":"Amidst the rapid transformation of the electricity supply and demand structure, there has been a consensus among policy makers on the need for further refinement of the time-of-use (ToU) pricing mechanism. Nonetheless, the challenge of capturing the dynamic interplay between ToU pricing design and the market behavior of diverse consumers, particularly in light of distributed energy integration, persists as an unresolved inquiry. This paper introduces a novel approach, utilizing reinforcement learning for the development of a ToU pricing model considering investment of distributed energy. The interaction between end consumers and utilities within the ToU pricing framework is encapsulated within a bi-level structure, characterized by significant applicability and scalability. A deep reinforcement learning algorithm is employed to train an agent in devising effective pricing strategies. To aid the agent in grasping the complex, interdependent pricing effects during peak, mid-peak, and valley periods, a configuration employing three interconnected Long Short-Term Memory networks is adopted. Case studies, grounded in empirical datasets, substantiate the efficacy and rationality of the methodology presented herein. It is anticipated that the framework proposed in this study will serve as a valuable reference for the design of efficient ToU pricing in diverse regions.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"12 2","pages":"997-1010"},"PeriodicalIF":6.7,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465713","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}
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：