Pub Date : 2025-12-15DOI: 10.1109/TNSE.2025.3644385
Zhen Yin;Wei Wang;Yanyu Cheng;Yiliang Liu;Xiaozhen Lu
The emerging increasingly sophisticated, intelligent, and stealthy network attacks pose more severe security threats to the edge network. In particular, the emergence of novel intelligent attacks makes it a challenging issue to obtain sufficient attack samples, and thus classical deep learning-driven intrusion detection frameworks (IDSs) become ineffective. To tackle this issue, we introduce a novel intrusion detection framework leveraging few-shot class-incremental learning (FSCIL) capabilities to achieve robust detection of emerging threats with few samples. This approach pre-trains a backbone traffic classification model and employs few-shot training with prototypical networks. To further reduce catastrophic forgetting while improving both accuracy and system robustness, we incrementally fine-tune the classification model with supervised contrastive learning, and also realize rapid adaptation to new attacks. Evaluations on the intrusion detection datasets CIC-IDS2017 and USTC-TF2016 demonstrate that our framework consistently outperforms baseline models for emerging attacks detection with few attack samples while preserving effective recognition of known threats.
{"title":"Efficient Intrusion Detection for Edge Network via Multi-Stage Few-Shot Class-Incremental Learning","authors":"Zhen Yin;Wei Wang;Yanyu Cheng;Yiliang Liu;Xiaozhen Lu","doi":"10.1109/TNSE.2025.3644385","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3644385","url":null,"abstract":"The emerging increasingly sophisticated, intelligent, and stealthy network attacks pose more severe security threats to the edge network. In particular, the emergence of novel intelligent attacks makes it a challenging issue to obtain sufficient attack samples, and thus classical deep learning-driven intrusion detection frameworks (IDSs) become ineffective. To tackle this issue, we introduce a novel intrusion detection framework leveraging few-shot class-incremental learning (FSCIL) capabilities to achieve robust detection of emerging threats with few samples. This approach pre-trains a backbone traffic classification model and employs few-shot training with prototypical networks. To further reduce catastrophic forgetting while improving both accuracy and system robustness, we incrementally fine-tune the classification model with supervised contrastive learning, and also realize rapid adaptation to new attacks. Evaluations on the intrusion detection datasets CIC-IDS2017 and USTC-TF2016 demonstrate that our framework consistently outperforms baseline models for emerging attacks detection with few attack samples while preserving effective recognition of known threats.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4689-4706"},"PeriodicalIF":7.9,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145830761","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 : 2025-12-15DOI: 10.1109/TNSE.2025.3643634
Kun Yan;Wenping Ma;Shaohui Sun
The 6G Computing Power Network (CPN) is envisioned to orchestrate vast, distributed computing resources for future intelligent applications. However, achieving efficient, trusted, and privacy-preserving computing resource sharing in this decentralized environment poses significant challenges. To address these intertwined issues, this article proposes a holistic blockchain and evolutionary algorithm-based computing resource sharing (BECS) mechanism. BECS is designed to dynamically and adaptively balance task offloading among computing resources within the 6G CPN, thereby enhancing resource utilization. We model computing resource sharing as a multi-objective optimization problem, aiming to navigate these trade-offs. To tackle this NP-hard problem, we devise a kernel-distance-based dominance relation and incorporate it into the Non-dominated Sorting Genetic Algorithm III (NSGA-III), thereby significantly enhancing population diversity. In addition, we propose a pseudonym scheme based on zero-knowledge proofs to protect user privacy during computing resource sharing. Finally, security analysis and simulation results demonstrate that BECS can effectively leverage all computing resources in the 6G CPN, thereby significantly improving resource utilization while preserving user privacy.
{"title":"BECS: A Privacy-Preserving Computing Resource Sharing Mechanism for 6G Computing Power Network","authors":"Kun Yan;Wenping Ma;Shaohui Sun","doi":"10.1109/TNSE.2025.3643634","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3643634","url":null,"abstract":"The 6G Computing Power Network (CPN) is envisioned to orchestrate vast, distributed computing resources for future intelligent applications. However, achieving efficient, trusted, and privacy-preserving computing resource sharing in this decentralized environment poses significant challenges. To address these intertwined issues, this article proposes a holistic blockchain and evolutionary algorithm-based computing resource sharing (BECS) mechanism. BECS is designed to dynamically and adaptively balance task offloading among computing resources within the 6G CPN, thereby enhancing resource utilization. We model computing resource sharing as a multi-objective optimization problem, aiming to navigate these trade-offs. To tackle this NP-hard problem, we devise a kernel-distance-based dominance relation and incorporate it into the Non-dominated Sorting Genetic Algorithm III (NSGA-III), thereby significantly enhancing population diversity. In addition, we propose a pseudonym scheme based on zero-knowledge proofs to protect user privacy during computing resource sharing. Finally, security analysis and simulation results demonstrate that BECS can effectively leverage all computing resources in the 6G CPN, thereby significantly improving resource utilization while preserving user privacy.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4725-4742"},"PeriodicalIF":7.9,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145830850","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}
Structural node embedding is a fundamental technique for encoding the topology of a graph into low-dimensional vectors. However, many existing methods generate position-dependent embeddings, meaning that structurally similar nodes are represented dissimilarly merely due to their distance in the graph. Furthermore, these approaches often lack interpretability and robustness against structural noise. To address these challenges, this paper introduces GraphQWalk, an interpretable, unsupervised, and position-independent method that leverages the continuous quantum walk to capture structural features. Inspired by quantum physics, GraphQWalk first computes initial node features from the average transition probabilities of a particle in a continuous quantum walk. These features, encoding multi-scale structural information, are then aggregated within multi-hop neighborhoods to incorporate local context. Extensive experiments demonstrate that GraphQWalk effectively captures diverse structural roles, achieving superior robustness and performance over baseline models in downstream tasks from classification to cross-graph alignment.
{"title":"GraphQWalk: Learning Structural Node Embeddings via Continuous Quantum Walk","authors":"Guojun Liu;Juanhong Zhao;Houzhou Wei;Zhengxiong Zhou;Yunfei Song;Xiaomei Zhou;Guangzhi Qi","doi":"10.1109/TNSE.2025.3644803","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3644803","url":null,"abstract":"Structural node embedding is a fundamental technique for encoding the topology of a graph into low-dimensional vectors. However, many existing methods generate position-dependent embeddings, meaning that structurally similar nodes are represented dissimilarly merely due to their distance in the graph. Furthermore, these approaches often lack interpretability and robustness against structural noise. To address these challenges, this paper introduces GraphQWalk, an interpretable, unsupervised, and position-independent method that leverages the continuous quantum walk to capture structural features. Inspired by quantum physics, GraphQWalk first computes initial node features from the average transition probabilities of a particle in a continuous quantum walk. These features, encoding multi-scale structural information, are then aggregated within multi-hop neighborhoods to incorporate local context. Extensive experiments demonstrate that GraphQWalk effectively captures diverse structural roles, achieving superior robustness and performance over baseline models in downstream tasks from classification to cross-graph alignment.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4779-4796"},"PeriodicalIF":7.9,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145886555","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 : 2025-12-11DOI: 10.1109/TNSE.2025.3642623
Siyu Qian;Xiaolin Wang;Fangfei Li;Yi'ang Ren
Edge computing enables real-time estimation for mission-critical Industrial Internet of Things (IIoT) systems by processing massive data streams at the network edge. Existing methods struggle to balance estimation performance and energy efficiency under varying reliability and latency requirements. We investigate Multi-Link Operation (MLO) to improve adaptability under dynamic wireless conditions. However, existing schemes offer limited flexibility in transmission modes and rely on simplified network models, which restricts the potential for MLO-specific performance optimization. To overcome these challenges, we propose a novel flexible transmission mode selection mechanism called Synchronous-Asynchronous Co-existence Multi-Link Aggregation (SACMLA). Based on switchable MLO modes, we formulate a scheduling-estimation co-design problem that optimizes the trade-off between estimation error covariance and transmission energy consumption. Given the environmental complexity and the action space explosion caused by the joint optimization of transmission mode, packet-to-link mapping, and energy allocation, we design a Proximal Policy Optimization (PPO)-based Hierarchical Deep Reinforcement Learning for Multi-Link Co-optimization (HDRL-MLC), a two-tier architecture where a PPO-based outer loop dynamically adjusts scheduling, and an inner loop optimizes energy allocation in MLO. Simulation results demonstrate the necessity of the SACMLA mechanism and highlight the superior performance of the PPO-based HDRL-MLC algorithm over flat algorithms in balancing multiple communication metrics and adapting to varying traffic conditions.
{"title":"Multi-Link Enabled Reliable and Low Latency Transmission Design for Edge Estimation","authors":"Siyu Qian;Xiaolin Wang;Fangfei Li;Yi'ang Ren","doi":"10.1109/TNSE.2025.3642623","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3642623","url":null,"abstract":"Edge computing enables real-time estimation for mission-critical Industrial Internet of Things (IIoT) systems by processing massive data streams at the network edge. Existing methods struggle to balance estimation performance and energy efficiency under varying reliability and latency requirements. We investigate Multi-Link Operation (MLO) to improve adaptability under dynamic wireless conditions. However, existing schemes offer limited flexibility in transmission modes and rely on simplified network models, which restricts the potential for MLO-specific performance optimization. To overcome these challenges, we propose a novel flexible transmission mode selection mechanism called Synchronous-Asynchronous Co-existence Multi-Link Aggregation (SACMLA). Based on switchable MLO modes, we formulate a scheduling-estimation co-design problem that optimizes the trade-off between estimation error covariance and transmission energy consumption. Given the environmental complexity and the action space explosion caused by the joint optimization of transmission mode, packet-to-link mapping, and energy allocation, we design a Proximal Policy Optimization (PPO)-based Hierarchical Deep Reinforcement Learning for Multi-Link Co-optimization (HDRL-MLC), a two-tier architecture where a PPO-based outer loop dynamically adjusts scheduling, and an inner loop optimizes energy allocation in MLO. Simulation results demonstrate the necessity of the SACMLA mechanism and highlight the superior performance of the PPO-based HDRL-MLC algorithm over flat algorithms in balancing multiple communication metrics and adapting to varying traffic conditions.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4707-4724"},"PeriodicalIF":7.9,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145830884","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 : 2025-12-10DOI: 10.1109/TNSE.2025.3642330
Kang Hao Cheong
In recent years, the adversarial robustness of deep neural networks has been widely studied. Several attack and defense models have been proposed to date for models focusing on balanced datasets. However, the adversarial robustness study for imbalanced (long-tailed) datasets has rarely been investigated. Recently, a study investigated the combination of re-balancing tricks focusing on adversarial robustness. However, such adversarial training may not be suitable for long-tailed robustness. We observe that the adversarial samples from long-tailed datasets require additional exploration divergence in the sampling space. To solve this problem, we propose a Rebalanced Divergence-Enhanced Adversarial Training method (RbDAT) to improve the adversarial robustness of long-tailed classes. Our model comprises two parts, a re-balanced loss to avoid discrimination for tail classes and adversarial distributional learning to fully explore the adversarial perturbation space. We demonstrate the effectiveness of the proposed method by outperforming the state-of-the-art (SOTA) method. i.e., RbDAT outperforms the SOTA method with a clear margin of 5.01% under a projected gradient descent (PGD) attack on the CIFAR-10-LT dataset.
{"title":"Rebalanced Divergence-Enhanced Adversarial Training for Long-Tailed Robustness","authors":"Kang Hao Cheong","doi":"10.1109/TNSE.2025.3642330","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3642330","url":null,"abstract":"In recent years, the adversarial robustness of deep neural networks has been widely studied. Several attack and defense models have been proposed to date for models focusing on balanced datasets. However, the adversarial robustness study for imbalanced (long-tailed) datasets has rarely been investigated. Recently, a study investigated the combination of re-balancing tricks focusing on adversarial robustness. However, such adversarial training may not be suitable for long-tailed robustness. We observe that the adversarial samples from long-tailed datasets require additional exploration divergence in the sampling space. To solve this problem, we propose a <bold>R</b>ebalanced <bold>D</b>ivergence-Enhanced <bold>A</b>dversarial <bold>T</b>raining method (RbDAT) to improve the adversarial robustness of long-tailed classes. Our model comprises two parts, a re-balanced loss to avoid discrimination for tail classes and adversarial distributional learning to fully explore the adversarial perturbation space. We demonstrate the effectiveness of the proposed method by outperforming the state-of-the-art (SOTA) method. i.e., RbDAT outperforms the SOTA method with a clear margin of 5.01% under a projected gradient descent (PGD) attack on the CIFAR-10-LT dataset.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"5721-5735"},"PeriodicalIF":7.9,"publicationDate":"2025-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146026349","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 : 2025-12-09DOI: 10.1109/TNSE.2025.3641579
Ye Wang;Jingjing Wang;Jianrui Chen;Xiangwang Hou;Ziyang Wang;Chunxiao Jiang
The evolution of the Internet of Vehicles (IoV) has introduced computation-intensive and latency-sensitive applications that challenge traditional cloud architectures. Although drone-aided IoV offers a flexible solution, it presents a complex optimization problem. The core challenge lies in balancing task offloading efficiency with crucial operational safety constraints, such as collision avoidance and battery management, a gap often overlooked in existing research. This paper addresses this problem by first modeling the drone-aided task offloading system as a constrained multi-agent Markov decision process. Based on this framework, we propose a novel safe multi-agent reinforcement learning algorithm (MARL) named Lagrangian-constrained multi-agent policy optimization (LC-MAPO). The LC-MAPO integrates safety constraints into the twin delayed deep deterministic policy gradient (TD3) actor-critic framework using Lagrangian duality theory. The algorithm's effectiveness was validated in three distinct simulation scenarios and compared against an unconstrained multi-agent deep deterministic policy gradient (MADDPG) algorithm and a greedy algorithm. Experimental results demonstrate that LC-MAPO achieves superior performance in both safety adherence and task processing efficiency.
{"title":"Drone-Aided Secure Task Offloading Optimization for Internet of Vehicles: Review, Challenges and Method","authors":"Ye Wang;Jingjing Wang;Jianrui Chen;Xiangwang Hou;Ziyang Wang;Chunxiao Jiang","doi":"10.1109/TNSE.2025.3641579","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3641579","url":null,"abstract":"The evolution of the Internet of Vehicles (IoV) has introduced computation-intensive and latency-sensitive applications that challenge traditional cloud architectures. Although drone-aided IoV offers a flexible solution, it presents a complex optimization problem. The core challenge lies in balancing task offloading efficiency with crucial operational safety constraints, such as collision avoidance and battery management, a gap often overlooked in existing research. This paper addresses this problem by first modeling the drone-aided task offloading system as a constrained multi-agent Markov decision process. Based on this framework, we propose a novel safe multi-agent reinforcement learning algorithm (MARL) named Lagrangian-constrained multi-agent policy optimization (LC-MAPO). The LC-MAPO integrates safety constraints into the twin delayed deep deterministic policy gradient (TD3) actor-critic framework using Lagrangian duality theory. The algorithm's effectiveness was validated in three distinct simulation scenarios and compared against an unconstrained multi-agent deep deterministic policy gradient (MADDPG) algorithm and a greedy algorithm. Experimental results demonstrate that LC-MAPO achieves superior performance in both safety adherence and task processing efficiency.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4596-4615"},"PeriodicalIF":7.9,"publicationDate":"2025-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145830813","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 : 2025-12-09DOI: 10.1109/TNSE.2025.3642184
Yixuan Lv;Lei Liu;Yan-Jun Liu;Yang Chen
In this article, based on nodes and edges, we research the fully distributed generalized Nash equilibrium (GNE) search problem for aggregate games under coupling constraints with bounded disturbances. In order to solve this problem, the node adaptive control law and edge adaptive control law are introdued. In the node-based adaptive control law, agents dynamically adjust their weights through parameters tuning, whereas the edge-based adaptive control law introduces additional adjustable parameters to enable the multi-agent system to autonomously regulate the gain size. Compared with the aggregation game discussed earlier, it can make the agent adjust its behavior under the condition of complete information. Then, it is proved that the GNE of the system can remain asymptotically stable under these two strategies. Although preliminary explorations of disturbances mitigation in GNE problems exist in the literature, the dynamic compensation mechanisms and system performance control under bounded disturbances in aggregation games with coupling constraints remain insufficiently explored. In this article, the disturbances compensations are designed to minimize the damage of the system by keeping the system stable when bounded disturbances occur in the aggregate game under coupling constraints. Finally, the real effectiveness of the designed strategy is proved by simulation experiments.
{"title":"Fully Distributed Nash Equilibrium Search for Aggregate Games Under Coupling Constraints With Bounded Disturbances","authors":"Yixuan Lv;Lei Liu;Yan-Jun Liu;Yang Chen","doi":"10.1109/TNSE.2025.3642184","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3642184","url":null,"abstract":"In this article, based on nodes and edges, we research the fully distributed generalized Nash equilibrium (GNE) search problem for aggregate games under coupling constraints with bounded disturbances. In order to solve this problem, the node adaptive control law and edge adaptive control law are introdued. In the node-based adaptive control law, agents dynamically adjust their weights through parameters tuning, whereas the edge-based adaptive control law introduces additional adjustable parameters to enable the multi-agent system to autonomously regulate the gain size. Compared with the aggregation game discussed earlier, it can make the agent adjust its behavior under the condition of complete information. Then, it is proved that the GNE of the system can remain asymptotically stable under these two strategies. Although preliminary explorations of disturbances mitigation in GNE problems exist in the literature, the dynamic compensation mechanisms and system performance control under bounded disturbances in aggregation games with coupling constraints remain insufficiently explored. In this article, the disturbances compensations are designed to minimize the damage of the system by keeping the system stable when bounded disturbances occur in the aggregate game under coupling constraints. Finally, the real effectiveness of the designed strategy is proved by simulation experiments.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4651-4668"},"PeriodicalIF":7.9,"publicationDate":"2025-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145830829","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}
Unmanned aerial vehicle-based mobile crowdsensing (UAV-MCS) has emerged as a promising paradigm for large-scale data collection in low-altitude environments. Efficient task allocation is critical in UAV-MCS to reduce the flight distance of UAVs and ensure timely data acquisition. However, since UAVs typically depart from the locations of their operators, task allocation based on geographical coordinates can leak the location privacy of both UAV operators (called workers for simplicity) and task requesters. Existing privacy-preserving task allocation schemes often focus solely on protecting the privacy of UAV operators, while neglecting that of task requesters, and many fail to achieve accurate allocation results, particularly in multi-task scenarios. To address these challenges, we propose an efficient and accurate privacy-preserving task allocation scheme for UAV-MCS that leverages additive secret sharing to simultaneously protect the locations of both UAV operators and task requesters. To enable accurate allocation, we design a privacy-preserving comparison protocol (PAC) based on additive secret sharing and an optimized Paillier cryptosystem. Moreover, to support multi-task allocation under privacy constraints, we develop a privacy-preserving version of the Hungarian method. Experimental results on both real-world and synthetic datasets demonstrate that our scheme effectively reduces UAV travel distance while preserving location privacy, outperforming existing schemes in both efficiency and accuracy.
{"title":"Efficient and Accurate Privacy-Preserving Task Allocation for Unmanned Aerial Vehicle-Based Mobile Crowdsensing","authors":"Bowen Zhao;Siyuan Guan;Minghui Chen;Jiali Wu;Cheng Qiao;Yang Xiao","doi":"10.1109/TNSE.2025.3641172","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3641172","url":null,"abstract":"Unmanned aerial vehicle-based mobile crowdsensing (UAV-MCS) has emerged as a promising paradigm for large-scale data collection in low-altitude environments. Efficient task allocation is critical in UAV-MCS to reduce the flight distance of UAVs and ensure timely data acquisition. However, since UAVs typically depart from the locations of their operators, task allocation based on geographical coordinates can leak the location privacy of both UAV operators (called workers for simplicity) and task requesters. Existing privacy-preserving task allocation schemes often focus solely on protecting the privacy of UAV operators, while neglecting that of task requesters, and many fail to achieve accurate allocation results, particularly in multi-task scenarios. To address these challenges, we propose an efficient and accurate privacy-preserving task allocation scheme for UAV-MCS that leverages additive secret sharing to simultaneously protect the locations of both UAV operators and task requesters. To enable accurate allocation, we design a privacy-preserving comparison protocol (PAC) based on additive secret sharing and an optimized Paillier cryptosystem. Moreover, to support multi-task allocation under privacy constraints, we develop a privacy-preserving version of the Hungarian method. Experimental results on both real-world and synthetic datasets demonstrate that our scheme effectively reduces UAV travel distance while preserving location privacy, outperforming existing schemes in both efficiency and accuracy.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"5638-5653"},"PeriodicalIF":7.9,"publicationDate":"2025-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146026407","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 this paper, a cell-free massive multiple-input multiple-output (CF-mMIMO) system coexisting with device-to-device (D2D) pairs is investigated under ultra-dense wireless networks (UDWNs), aiming to accommodate the growing number of devices and escalating data traffic demands in future sixth-generation (6G) networks. To overcome the challenge of pilot shortage in UDWNs, non-orthogonal multiple-access (NOMA) and successive interference cancellation (SIC) technologies are employed, enabling users within the same cluster to share pilots efficiently. A closed-form expression for the spectral efficiency of conjugate beamforming receivers is derived, revealing that pilot contamination caused by pilot sharing severely degrades system performance. A novel hypergraph coloring (HGC)-based pilot assignment algorithm is proposed to address this issue, effectively capturing the complex cumulative interference among multiple users and D2D pairs. The algorithm constructs users’ interference hypergraphs using a user-centric access point (AP) selection strategy and allocates pilots based on interference weights within these hyperedges. Numerical results demonstrate that the proposed scheme significantly improves spectral efficiency, offering a promising solution for interference management in CF-mMIMO systems under UDWNs.
{"title":"Efficient Pilot Assignment for D2D-Underlaid NOMA Cell-Free Massive MIMO Systems: A Hypergraph Coloring Approach","authors":"Qin Wang;Haotian Chang;Yongxu Zhu;Mei Chen;Li Gao;Wenchao Xia;Haitao Zhao","doi":"10.1109/TNSE.2025.3640946","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3640946","url":null,"abstract":"In this paper, a cell-free massive multiple-input multiple-output (CF-mMIMO) system coexisting with device-to-device (D2D) pairs is investigated under ultra-dense wireless networks (UDWNs), aiming to accommodate the growing number of devices and escalating data traffic demands in future sixth-generation (6G) networks. To overcome the challenge of pilot shortage in UDWNs, non-orthogonal multiple-access (NOMA) and successive interference cancellation (SIC) technologies are employed, enabling users within the same cluster to share pilots efficiently. A closed-form expression for the spectral efficiency of conjugate beamforming receivers is derived, revealing that pilot contamination caused by pilot sharing severely degrades system performance. A novel hypergraph coloring (HGC)-based pilot assignment algorithm is proposed to address this issue, effectively capturing the complex cumulative interference among multiple users and D2D pairs. The algorithm constructs users’ interference hypergraphs using a user-centric access point (AP) selection strategy and allocates pilots based on interference weights within these hyperedges. Numerical results demonstrate that the proposed scheme significantly improves spectral efficiency, offering a promising solution for interference management in CF-mMIMO systems under UDWNs.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4636-4650"},"PeriodicalIF":7.9,"publicationDate":"2025-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145830799","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}
With the proliferation of Internet of Things (IoT) devices, IoT systems face increasingly severe threats from large-scale cyberattacks during network communications. To effectively recognize potential network threats and ensure the security of IoT, network traffic anomaly detection has been widely studied. Due to the high cost of manual labeling in real-world scenarios, only limited network traffic data is explicitly labeled as abnormal or normal. However, most existing anomaly detection methods struggle to enhance detection accuracy with minimal supervision and perform ineffectively at identifying potential unknown anomalies in unlabeled data. To address these limitations, this paper proposes WADE, a Weakly supervised Anomaly DEtection method based on Deep Reinforcement Learning (DRL). WADE enables simultaneous identification of known and unknown anomaly traffic with limited labeled data. Specifically, it incorporates a Dueling Q-network architecture and introduces a novel reward optimization mechanism that: 1) strengthens feature extraction from unlabeled data and 2) elevates decision accuracy via dynamic reward adaptation. Extensive experiments on IoT traffic datasets demonstrate that WADE outperforms four peer weakly supervised methods, achieving performance improvements of 12.2$%$ in Area Under the Precision-Recall Curve (AUC-PR) and 6.6$%$ in Area Under the Receiver Operating Characteristic Curve (AUC-ROC), which validate the effectiveness of WADE in safeguarding IoT systems.
{"title":"DRL-Based Weakly Supervised Traffic Anomaly Detection for IoT Networks","authors":"Ziteng Wang;Yiheng Ruan;Xianjun Deng;Xiaoxuan Fan;Shenghao Liu;Wei Feng;Deng Zhang","doi":"10.1109/TNSE.2025.3640175","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3640175","url":null,"abstract":"With the proliferation of Internet of Things (IoT) devices, IoT systems face increasingly severe threats from large-scale cyberattacks during network communications. To effectively recognize potential network threats and ensure the security of IoT, network traffic anomaly detection has been widely studied. Due to the high cost of manual labeling in real-world scenarios, only limited network traffic data is explicitly labeled as abnormal or normal. However, most existing anomaly detection methods struggle to enhance detection accuracy with minimal supervision and perform ineffectively at identifying potential unknown anomalies in unlabeled data. To address these limitations, this paper proposes WADE, a <italic>Weakly supervised Anomaly DEtection</i> method based on Deep Reinforcement Learning (DRL). WADE enables simultaneous identification of known and unknown anomaly traffic with limited labeled data. Specifically, it incorporates a Dueling Q-network architecture and introduces a novel reward optimization mechanism that: 1) strengthens feature extraction from unlabeled data and 2) elevates decision accuracy via dynamic reward adaptation. Extensive experiments on IoT traffic datasets demonstrate that WADE outperforms four peer weakly supervised methods, achieving performance improvements of 12.2<inline-formula><tex-math>$%$</tex-math></inline-formula> in Area Under the Precision-Recall Curve (AUC-PR) and 6.6<inline-formula><tex-math>$%$</tex-math></inline-formula> in Area Under the Receiver Operating Characteristic Curve (AUC-ROC), which validate the effectiveness of WADE in safeguarding IoT systems.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4562-4577"},"PeriodicalIF":7.9,"publicationDate":"2025-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145879968","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：