IEEE Transactions on Network Science and Engineering最新文献_第10页

Split Learning Based Cloud-Edge-End Collaborative Model Training in Heterogeneous Networks 异构网络中基于分裂学习的云-端协同模型训练

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2025-08-08 DOI: 10.1109/TNSE.2025.3597161

Jian Wang;Gang Feng;Yi-Jing Liu;Xinyi Xu;Lei Cheng;Wei Jiang;Li Ping Qian

{"title":"Split Learning Based Cloud-Edge-End Collaborative Model Training in Heterogeneous Networks","authors":"Jian Wang;Gang Feng;Yi-Jing Liu;Xinyi Xu;Lei Cheng;Wei Jiang;Li Ping Qian","doi":"10.1109/TNSE.2025.3597161","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3597161","url":null,"abstract":"Large language models (LLMs) demonstrate significant potential for enabling intelligent endogenous networks owing to their amazing intelligence level. LLMs are currently deployed on cloud servers as their vast parameter scales introduce a substantial computational burden to pre-training and inference processes. This deployment paradigm faces severe challenges, including insufficient personalization, high inference latency, and privacy concerns, as an increasing number of mobile users enjoy the LLM services. To enhance the personalization of pre-trained LLMs, exploiting massive local datasets distributed across edge devices to fine-tune LLMs is essential. In this paper, we propose a split learning-based cloud-edge-end collaborative training framework (SCCT) to harness the abundant computational resources of cloud and edge servers while exploiting massive distributed datasets on edge devices for LLM fine-tuning. In SCCT, an LLM with a parameter-efficient fine-tuning module is deployed on the cloud server, while a small-scale language model (SLM) is deployed across each edge device participating in SCCT and its associated edge server in the manner of split learning. SLM and LLM collaborate in a serial manner for training and inference. To optimize the collaboration efficiency, we formulate a mixed integer nonlinear programming problem to minimize the training latency of SCCT, considering the resource heterogeneity of edge devices and network dynamics. To solve this problem, a two-timescale optimization algorithm is proposed to determine the optimal split points on the large timescale and the allocation of communication and computational resources of the edge server on the small timescale. Numerical results demonstrate that the proposed two-timescale optimization algorithm assisted SCCT outperforms the state-of-the-art LLMs deployment and training paradigm in terms of training efficiency and inference performance.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"1569-1585"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148514028","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}

引用次数: 0

Joint Optimization of Data Placement and Task Scheduling for Distributed Green Data Centers With Optical Satellite Networks 基于光卫星网络的分布式绿色数据中心数据布局与任务调度联合优化

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2026-03-11 DOI: 10.1109/TNSE.2026.3672902

Hiroto Oshima;Hiroaki Hashida;Yuichi Kawamoto;Nei Kato;Kazushi Sugyo;Yohei Hasegawa;Masayuki Ariyoshi

{"title":"Joint Optimization of Data Placement and Task Scheduling for Distributed Green Data Centers With Optical Satellite Networks","authors":"Hiroto Oshima;Hiroaki Hashida;Yuichi Kawamoto;Nei Kato;Kazushi Sugyo;Yohei Hasegawa;Masayuki Ariyoshi","doi":"10.1109/TNSE.2026.3672902","DOIUrl":"https://doi.org/10.1109/TNSE.2026.3672902","url":null,"abstract":"Distributed green data centers powered by renewable energy have garnered considerable attention owing to their potential to enhance both sustainability and disaster resilience. To interconnect these geographically dispersed sites, low-Earth orbit satellite optical communications offer a promising solution, providing low-latency and high-bandwidth connectivity. However, the efficient operation of such a system hinges on a comprehensive task scheduling strategy that addresses two critical factors: task assignment—determining which data center should process each task—and processor performance control—regulating the operational speed of each processor. The effectiveness of this strategy is deeply intertwined with data placement decisions, as the location of stored data directly influences the set of data centers eligible to execute specific tasks. Optimizing this system is particularly challenging owing to the inherent variability in two primary factors: the availability of renewable energy and the quality of satellite communication links, both of which fluctuate with weather. To address these challenges, this study proposes a framework for the joint optimization of data placement and task scheduling. The proposed approach simultaneously considers fluctuating communication and power resources to maximize user task satisfaction. The problem is formulated as a mixed-integer nonlinear programming problem, and efficient solutions are obtained through computationally efficient alternating optimization algorithms. The proposed method significantly enhances task satisfaction rates compared with conventional approaches that optimize data placement and task scheduling in isolation, particularly under heterogeneous task distributions and adverse weather conditions, demonstrating its robustness and adaptability.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"7565-7580"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148514161","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}

引用次数: 0

LBKD: Rethinking Federated Backdoors for Low-Altitude Economy via LLMs and Bidirectional Knowledge Distillation LBKD：基于llm和双向知识蒸馏的低空经济联邦后门再思考

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2025-10-28 DOI: 10.1109/TNSE.2025.3626056

Haoquan Wang;Boyuan Li;Panpan Chen;Libing Wu;Zijian Li;Tony Q.S. Quek

{"title":"LBKD: Rethinking Federated Backdoors for Low-Altitude Economy via LLMs and Bidirectional Knowledge Distillation","authors":"Haoquan Wang;Boyuan Li;Panpan Chen;Libing Wu;Zijian Li;Tony Q.S. Quek","doi":"10.1109/TNSE.2025.3626056","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3626056","url":null,"abstract":"With the growing adoption of low-altitude economy (LAE), such as urban air mobility and smart logistics, federated learning (FL) has become a key technology for enabling distributed intelligence while preserving data privacy. However, edge clients in such systems often suffer from limited resources and heterogeneous data, making them especially vulnerable to covert backdoor attacks. These threats can compromise critical functions such as navigation control or data confidentiality. Existing backdoor methods typically target specific clients, but their performance significantly degrades on edge clients due to insufficient data and computation power. To address this challenge, we propose LBKD, a novel backdoor attack strategy that combines large language models (LLMs) with bidirectional knowledge distillation. In our method, the attacker first obtains the target client's model and data, then uses LLMs guided by poisoning and data generation instructions to produce both poisoned and synthetic samples. During training, a more expressive distillation model engages in two-way knowledge exchange with the client model, enhancing its ability to learn and retain the backdoor pattern. The final poisoned model is submitted to the server, successfully embedding the backdoor into the global model. Experimental results show that <bold>LBKD</b> significantly improves attack success rates (ASR), resulting in up to a 75.4% improvement over existing methods. Even under defense mechanisms, the ASR remains above 90% in some cases, demonstrating the method's effectiveness and robustness on edge clients.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4422-4439"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148514381","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}

引用次数: 0

Harnessing the Power of LLMs, Informers and Decision Transformers for Intent-Driven RAN Management in 6G 利用llm、信息者和决策转换器的力量实现6G中意图驱动的RAN管理

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2025-08-05 DOI: 10.1109/TNSE.2025.3596028

Md Arafat Habib;Pedro Enrique Iturria-Rivera;Yigit Ozcan;Medhat Elsayed;Majid Bavand;Raimundas Gaigalas;Melike Erol-Kantarci

{"title":"Harnessing the Power of LLMs, Informers and Decision Transformers for Intent-Driven RAN Management in 6G","authors":"Md Arafat Habib;Pedro Enrique Iturria-Rivera;Yigit Ozcan;Medhat Elsayed;Majid Bavand;Raimundas Gaigalas;Melike Erol-Kantarci","doi":"10.1109/TNSE.2025.3596028","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3596028","url":null,"abstract":"Intent-driven network management is critical for managing the complexity of 5G and 6G networks. It enables adaptive, on-demand management of the network based on the objectives of the network operators. In this paper, we propose an innovative three-step framework for intent-driven network management based on Generative AI (GenAI) algorithms. First, we fine-tune a Large Language Model (LLM) on a custom dataset using a Quantized Low-Rank Adapter (QLoRA) to enable memory-efficient intent processing within limited computational resources. A Retrieval Augmented Generation (RAG) module is included to support dynamic decision-making. Second, we utilize a transformer architecture for time series forecasting to predict key parameters, such as power consumption, traffic load, and packet drop rate, to facilitate intent validation proactively. Lastly, we introduce a Hierarchical Decision Transformer with Goal Awareness (HDTGA) to optimize the selection and orchestration of network apPlications in Radio Access Network (RAN). Our intent guidance and processing approach improves BERTScore by 6%, and the semantic similarity score by 9% compared to the base LLM model. Again, the proposed predictive intent validation approach can successfully rule out the performance-degrading intents with an average of 88% accuracy. Finally, compared to the baselines, the proposed HDTGA algorithm improves RAN performance by increasing throughput by at least 19.3%, reducing delay by 48.5%, and boosting energy efficiency by 54.9%.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4187-4206"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148514360","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}

引用次数: 0

Volatile Event-Triggered Control for Stochastic Hybrid Networks Vulnerable to DoS Attacks 易受DoS攻击的随机混合网络易失性事件触发控制

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2026-02-06 DOI: 10.1109/TNSE.2026.3661985

Dongsheng Xu;Jianju Xu;Choon Ki Ahn;Huan Su

{"title":"Volatile Event-Triggered Control for Stochastic Hybrid Networks Vulnerable to DoS Attacks","authors":"Dongsheng Xu;Jianju Xu;Choon Ki Ahn;Huan Su","doi":"10.1109/TNSE.2026.3661985","DOIUrl":"https://doi.org/10.1109/TNSE.2026.3661985","url":null,"abstract":"In this paper, the stability of stochastic complex networks with impulsive effects vulnerable to asynchronous DoS attacks is studied. The volatile event-triggered control (ETC) is advanced, featuring a control gain characterized by a sign-indefinite function that follows specific volatile patterns. Furthermore, the concept of average volatile gain under DoS attacks is proposed to quantify the volatile control gain. To address the challenges posed by volatile control gain and DoS attacks, a generalized differential inequality is proposed, resulting in a broader application of the main results. Using stochastic analysis techniques, graph-theoretic methodology, and the Lyapunov method, improved stability criteria are established. In particular, the joint potential unstable factors of impulsive effects and DoS attacks on stability are theoretically revealed. Stochastic hybrid networks that apply ETC can achieve stability under certain conditions if the magnitude of the impulsive effects and the frequency and duration of DoS attacks are sufficiently limited. Additionally, by incorporating the dwell time into ETC, the Zeno phenomenon can be avoided. Finally, a numerical example is provided to illustrate the validity of the obtained results.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"6655-6668"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148514495","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}

引用次数: 0

Slicing Resource Scheduling for 6G Green Cellular-Vehicular Networks: A Secrecy Oriented Energy-Efficient Approach 6G绿色蜂窝车联网的切片资源调度：一种面向保密的节能方法

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2025-06-19 DOI: 10.1109/TNSE.2025.3581201

Minghui Dai;Tianshun Wang;Liping Qian;Zhou Su;Yuan Wu

{"title":"Slicing Resource Scheduling for 6G Green Cellular-Vehicular Networks: A Secrecy Oriented Energy-Efficient Approach","authors":"Minghui Dai;Tianshun Wang;Liping Qian;Zhou Su;Yuan Wu","doi":"10.1109/TNSE.2025.3581201","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3581201","url":null,"abstract":"The 6G empowered Internet of connected vehicles (IoCVs) have paved the way to the autonomous driving era, where the ultra-low latency communication and ultra-reliable connections promote the quality of experience (QoE) for vehicle users. However, the high data traffic load and communication resource constraint pose a heavy burden to autonomous driving. Moreover, the energy consumption and secrecy data transmission are critical issues that influence the service efficiency of 6G green vehicular networks. To address the above challenges, in this paper, we present a secrecy oriented energy-efficient slicing resource scheduling framework in 6G green cellular-vehicular networks, with the objective of promoting the service efficiency for vehicle users. In specific, the IoCVs can connect to roadside units (RSUs) through vehicle-to-infrastructure links. We consider that non-internet of connected vehicles (N-IoCVs) and eavesdroppers coexist in the networks, where N-IoCVs can reuse the resource block non-orthogonally with IoCVs for information exchange, and the eavesdroppers may overhear the data transmission of IoCVs and N-IoCVs. To guarantee the requirement of quality of service and meet the green communication in autonomous driving, we design a joint energy efficiency and slicing resource scheduling problem subject to the data rate and secrecy capacity requirements. To address the formulated non-convex optimization problem, we transform the resource scheduling problem into an equivalent form, and exploit a vertical decomposition approach to solve it. Regarding each decomposed subproblem, we propose the corresponding algorithms to yield the optimal strategy for slicing resource scheduling. Finally, numerical results show the advantages of our proposed slicing resource scheduling scheme for vehicular networks compared to several benchmark algorithms.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"67-83"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148514778","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}

引用次数: 0

GCB-PPO2: A Hybrid Deep Reinforcement Learning Intrusion Detection System for Under-Represented Attack Categories in SDN GCB-PPO2：针对SDN中未充分代表攻击类别的混合深度强化学习入侵检测系统

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2025-06-20 DOI: 10.1109/TNSE.2025.3581689

Chen Jue;Tao Hongyu;Cui Meng;Peng Haidong;Qiu Xihe

{"title":"GCB-PPO2: A Hybrid Deep Reinforcement Learning Intrusion Detection System for Under-Represented Attack Categories in SDN","authors":"Chen Jue;Tao Hongyu;Cui Meng;Peng Haidong;Qiu Xihe","doi":"10.1109/TNSE.2025.3581689","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3581689","url":null,"abstract":"The centralized control plane inherent in SDN architecture creates critical security dependencies, where malicious exploitation of controller vulnerabilities could propagate systemic network failures. The Intrusion Detection System (IDS) effectively counters cyber threats, and Deep Reinforcement Learning (DRL) can enable the IDS to dynamically adapt to the constantly evolving attack patterns through autonomous environmental interaction and real-time policy optimization. However, the current DRL based IDS face the limitations of minority-class detection, generalization gap, and tuning complexity. This paper propose GCB-PPO2 (Generative Adversarial Networks CNNA-BiLSTM Proximal Policy Optimization 2), a hybrid DRL system with the following three innovations: (1) We design a C-GAN (Conditional Generative Adversarial Networks) architecture to generate targeted under-represented attack samples for the SDN scenario intentionally, improving the recognition accuracy of the model for minority classes significantly. (2) We propose a PPO2-based DRL framework with a CNN-LSTM shared network to optimize dynamic policy adaptation while capturing spatio-temporal patterns, enhancing generalization ability to different environmental dynamics. (3) We embed a Bayesian optimization as a dedicated hyper-parameter auto-tuner, systematically resolving sensitivity bottlenecks and ensuring robustness in fluctuating environments. Experimental results demonstrate that GCB-PPO2 model achieves remarkable accuracy of 99.92% and 99.01% for binary and multiple classification scenarios, respectively, with F1-scores exceeding 85.91% for under-represented attack categories in InSDN dataset. Moreover, the model proves its strong generalization capability for maintaining over 98.65% accuracy on another dataset, while confirming its improved training efficiency for surpassing conventional hybrid deep learning models by more than 22.27% in training time.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"84-101"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148514737","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}

引用次数: 0

BCC-SRN: Blockchain-Based Cooperative Communication Scheme for Symbiotic Radio Network BCC-SRN：基于区块链的共生无线网络协同通信方案

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2025-09-26 DOI: 10.1109/TNSE.2025.3614744

Zhen Gao;Shipeng Lin;Zezhong Li;Jiuzhi Zhang;Zhaohui Guo;Zhu Han;Xiaofei Wang;Ying-chang Liang

{"title":"BCC-SRN: Blockchain-Based Cooperative Communication Scheme for Symbiotic Radio Network","authors":"Zhen Gao;Shipeng Lin;Zezhong Li;Jiuzhi Zhang;Zhaohui Guo;Zhu Han;Xiaofei Wang;Ying-chang Liang","doi":"10.1109/TNSE.2025.3614744","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3614744","url":null,"abstract":"The symbiotic radio network (SRN) plays a crucial role in green communication solutions by facilitating cooperative communication between devices of primary and secondary links, leading to enhanced transmission quality and optimized resource efficiency. For the joint demodulation of primary and secondary signals, the multiparty trust is essential for operators with different interests. Building multiparty trust in a decentralized manner is challenging for the implementation of the SRN. As an emerging distributed ledger technology, blockchain provides multiparty trust between unreliable devices to share data and resources cooperatively. This paper proposes a consortium Blockchain-based Cooperative Communication scheme for SRN (BCC-SRN), where Hyperledger Fabric is used as the trust anchor with better security, higher throughput, and more flexible organization. In particular, a credit-based incentive mechanism is established to encourage different parties to behave according to the negotiated contracts. The credit is recorded by smart contracts and adjusted based on the contribution of the corresponding cooperators. Simulation results prove that the proposed BCC-SRN is highly effective in improving cooperation security and transmission quality.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"2371-2387"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148515121","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}

引用次数: 0

Location-Aware Anti-Jamming Game: An Intelligent Approach for Reliable Wireless Communications 位置感知抗干扰博弈：一种可靠无线通信的智能方法

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2025-10-06 DOI: 10.1109/TNSE.2025.3617848

Chengxi Li;Wei Gong;Minghui Liwang;Li Li;Baoxian Zhang;Cheng Li;Jie Chen

{"title":"Location-Aware Anti-Jamming Game: An Intelligent Approach for Reliable Wireless Communications","authors":"Chengxi Li;Wei Gong;Minghui Liwang;Li Li;Baoxian Zhang;Cheng Li;Jie Chen","doi":"10.1109/TNSE.2025.3617848","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3617848","url":null,"abstract":"Advances in wireless communication technologies have led to increasing dynamic jamming attacks. Enabling the receiver to make real-time decisions that enhance the communication quality from the transmitter and to defend against the interference from the jammer becomes a key challenge, which can be modeled as a game. In such a game, the receiver aims to optimize communication performance by adjusting its geographical location and negotiating frequency band. Meanwhile, the jammer tries to launch more effective attacks based on the receiver's decisions. Nevertheless, the complex electromagnetic environment, combined with the vast decision possibilities thereby induced, poses significant challenges in accurately quantifying payoff matrices within such games. Specifically, two-dimensional geographical location and one-dimensional frequency band form a three-dimensional strategy space, further introducing difficulties for intelligent and efficient decision-making. To address these challenges, we first develop a location-aware anti-jamming game (LAJ game) to characterize the simultaneous decision-making and non-cooperative relationship between the receiver and the jammer, with Nash Equilibrium (NE) being the optimal strategy for both game players. Through appropriate discretization, we then compute two payoff matrices based on geographical locations and frequency bands. Finally, we propose ResBiNet, a continuous series-parallel architecture that utilizes inverted residual blocks to construct a deep cascade residual network, specifically designed to effectively solve large-scale LAJ games. Extensive experiments using real-world and synthesized datasets demonstrate that our proposed method can exponentially reduce running time while achieving an approximation error of less than 5%, offering an intelligent and efficient solution to address location-aware anti-jamming communication challenges.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"2486-2502"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148515412","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}

引用次数: 0

Cooperative Multi-Agent Deep Reinforcement Learning for Service-Aware Energy Optimization at the Network Edge 面向网络边缘服务感知能量优化的协同多智能体深度强化学习

IF 7.3 2区计算机科学

IEEE Transactions on Network Science and Engineering Pub Date : 2026-01-01 Epub Date: 2025-11-17 DOI: 10.1109/TNSE.2025.3633581

Shreya K. Chari;John S. Vardakas;Kostas Ramantas;Christos Verikoukis

{"title":"Cooperative Multi-Agent Deep Reinforcement Learning for Service-Aware Energy Optimization at the Network Edge","authors":"Shreya K. Chari;John S. Vardakas;Kostas Ramantas;Christos Verikoukis","doi":"10.1109/TNSE.2025.3633581","DOIUrl":"https://doi.org/10.1109/TNSE.2025.3633581","url":null,"abstract":"Colossal expansion of network services has increased the need for faster and efficient allocation of network resources to user applications. Moving forward, the energy generated at the network edge will impact the offloading of tasks. A Multi-access Edge Computing (MEC) unit now faces the challenging task of optimizing the network resource orchestration while maintaining sustainable energy emissions. In this paper, an application-aware cooperative Multi-Agent Deep Reinforcement Learning (MADRL) algorithm is proposed to mitigate the problem of energy and delay management at the MEC. Each agent within the system uses the Soft-Actor Critic (SAC) as the learning policy while maintaining an overall approach of centralized training and decentralized execution. Network delay and energy consumption using CPU and memory resources in the MEC are used as performance indicators. The performance of the MADRL algorithm is measured against that of a single-agent Actor-Critic (AC) algorithm, Multi-Agent Proximal Policy Optimization (MAPPO) and Monte-Carlo Simulations (MCS). The results illustrate a significantly finer performance by the proposed MADRL algorithm leading to reduced energy consumption and lower delay values. Further, the need for application-aware resource management for green computing in rapidly expandable networks in the future is established.","PeriodicalId":54229,"journal":{"name":"IEEE Transactions on Network Science and Engineering","volume":"13 ","pages":"4124-4137"},"PeriodicalIF":7.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148515413","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}

引用次数: 0