IEEE Transactions on Network and Service Management最新文献

{"title":"Joint DNN Partitioning and Task Offloading Based on Attention Mechanism-Aided Reinforcement Learning","authors":"Mengyuan Zhang;Juan Fang;Ziyi Teng;Yaqi Liu;Shen Wu","doi":"10.1109/TNSM.2025.3561739","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3561739","url":null,"abstract":"The rapid advancement of artificial intelligence applications has resulted in the deployment of a growing number of deep neural networks (DNNs) on mobile devices. Given the limited computational capabilities and small battery capacity of these devices, supporting efficient DNN inference presents a significant challenge. This paper considers the joint design of DNN model partitioning and offloading under high-concurrent tasks scenarios. The primary objective is to accelerate DNN task inference and reduce computational delay. Firstly, we propose an innovative adaptive inference framework that partitions DNN models into interdependent sub-tasks through a hierarchical partitioning method. Secondly, we develop a delay prediction model based on a Random Forest (RF) regression algorithm to estimate the computational delay of each sub-task on different devices. Finally, we designed a high-performance DNN partitioning and task offloading method based on an attention mechanism-aided Soft Actor-Critic (AMSAC) algorithm. The bandwidth allocation for each user is determined by the attention mechanism based on the characteristics of the DNN tasks, and the Soft Actor-Critic algorithm is used for adaptive layer-level partitioning and offloading of the DNN model, reducing collaborative inference delay. Extensive experiments demonstrate that our proposed AMSAC algorithm effectively reduces DNN task inference latency cost and improves service quality.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2914-2927"},"PeriodicalIF":4.7,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10969114","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144232002","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Joint Power Allocation and Task Scheduling for Data Offloading in Non-Geostationary Orbit Satellite Networks","authors":"Lijun He;Ziye Jia;Juncheng Wang;Erick Lansard;Zhu Han;Chau Yuen","doi":"10.1109/TNSM.2025.3561266","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3561266","url":null,"abstract":"In Non-Geostationary Orbit Satellite Networks (NGOSNs) with a large number of battery-carrying satellites, proper power allocation and task scheduling are crucial to improving data offloading efficiency. In this work, we jointly optimize power allocation and task scheduling to achieve energy-efficient data offloading in NGOSNs. Our goal is to properly balance the minimization of the total energy consumption and the maximization of the sum weights of tasks. Due to the tight coupling between power allocation and task scheduling, we first derive the optimal power allocation solution to the joint optimization problem with any given task scheduling policy. We then leverage the conflict graph model to transform the joint optimization problem into an Integer Linear Programming (ILP) problem with any given power allocation strategy. We explore the unique structure of the ILP problem to derive an efficient semidefinite relaxation-based solution. Finally, we utilize the genetic framework to combine the above special solutions as a two-layer solution for the original joint optimization problem. Simulation results demonstrate that our proposed solution can properly balance the reduction of total energy consumption and the improvement of the sum weights of tasks, thus achieving superior system performance over the current literature.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2882-2896"},"PeriodicalIF":4.7,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144231990","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}

{"title":"Energy-Efficient Node Localization in Time-Varying UAV-RIS-Assisted and Cluster-Based IoT Networks","authors":"Vikash Kumar Bhardwaj;Aagat Shukla;Om Jee Pandey","doi":"10.1109/TNSM.2025.3561269","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3561269","url":null,"abstract":"This paper proposes a novel method for energy-efficient node localization in time-varying Internet of Things (IoT) networks. The method utilizes Unmanned Aerial Vehicles (UAVs) and Reconfigurable Intelligent Surfaces (RISs) over cluster-based IoT networks, resulting in improved localization accuracy and Signal-to-Interference plus Noise Ratio (SINR) at the Base Station (BS). First, the proposed method computes the approximate coordinates of the User Equipments (UEs) through trilateration, utilizing a dataset comprising the coordinates of anchor nodes and Received Signal Strength (RSS) between UE-RIS pairs. Subsequently, K-means clustering is applied to efficiently group UEs based on their spatial proximity, leading to optimal RIS requirements. To further enhance the localization precision of the UEs, a Reinforcement Learning (RL) algorithm with a collision avoidance mechanism is employed over UAVs mounted with RIS. This innovative approach dynamically relocates a UAV-RIS pair to a maximum SINR position over the cluster. To compute the SINR value over a spatial location in the network, a novel approach is proposed herein, which utilizes a radio map of the network. Subsequently, the relocation of the UAV-RIS pair is followed by a novel method for computing the optimal phases of RIS elements, maximizing SINR at the BS. The final step involves Capon beamforming, strategically applied to antenna elements at the BS, resulting in further SINR improvement at the BS. The holistic integration of trilateration, clustering, RL, and beamforming collectively contributes to a system that achieves energy-efficiency, accurate localization, and enhanced SINR at BS. Experimental results demonstrate the effectiveness of the proposed methods, showcasing their potential for application in real-world scenarios where energy consumption and localization accuracy are critical considerations. To validate the significance of the proposed methods’ utilization, the proposed methods’ performance is also compared with that of existing methods.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2897-2913"},"PeriodicalIF":4.7,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144232173","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}

{"title":"MOHFL: Multi-Level One-Shot Hierarchical Federated Learning With Enhanced Model Aggregation Over Non-IID Data","authors":"Huili Liu;Yinglong Ma;Chenqi Guo;Xiaofeng Liu;Tingdong Wang","doi":"10.1109/TNSM.2025.3560629","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3560629","url":null,"abstract":"Hierarchical federated learning (HFL) is a privacy-preserving distributed machine learning framework with a client-edge-cloud hierarchy, where multiple edge servers perform partial model aggregation to reduce costly communication with the cloud server. Nevertheless, most existing HFL methods require extensive iterative communication and public datasets, which not only increase communication overhead but also raise privacy and security concerns. Moreover, non-independent and identically distributed (non-IID) data among devices can significantly impact the accuracy of the global model in HFL. To address these challenges, we propose a multi-level one-shot HFL framework (MOHFL), which aims to improve the performance of the global model in a single communication round. Specifically, we employ conditional variational autoencoders (CVAEs) as local models and use the aggregated decoders to generate an IID training set for the global model, thereby mitigating the negative impact of non-IID data. We improve the performance of CVAEs under different levels of data heterogeneity through a dominant class-based data selection method. Subsequently, an edge aggregation scheme based on multi-teacher knowledge distillation and contrastive learning is proposed to aggregate the knowledge from local decoders to edge decoders. Extensive experiments on four real-world datasets demonstrate that MOHFL is very competitive against four state-of-the-art baselines under various settings.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2853-2865"},"PeriodicalIF":4.7,"publicationDate":"2025-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144232188","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}

{"title":"Asynchronous Federated Caching Strategy for Multi-Satellite Collaboration Based on Deep Reinforcement Learning","authors":"Min Jia;Liang Zhang;Jian Wu;Qing Guo;Xuemai Gu","doi":"10.1109/TNSM.2025.3560833","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3560833","url":null,"abstract":"By incorporating caching functions into Low Earth Orbit (LEO) satellites, users worldwide can benefit from caching services. However, satellite caching faces the following challenges: 1) The continuous mobility of satellites introduces dynamic shifts in user distribution, resulting in unpredictable variations in interested content over time. 2) The cached content is susceptible to becoming obsolete due to the brief connection times established between satellites and clients. 3) Significant concerns arise regarding data privacy and security. Users may exhibit reluctance to transmit local data for privacy protection. To address the abovementioned challenges, we propose an asynchronous federated caching strategy (AFCS) consisting of an access satellite and several collaboration satellites. Clients employ an asynchronous federated learning methodology to collaboratively train a global model for predicting content popularity. Concerning privacy protection, clients are not required to upload local data. Instead, they only need to transmit the model hyperparameters. This approach significantly diminishes the risk of data leakage, thereby safeguarding data privacy effectively. We propose a novel strategy for client selection participating in global model training. Through model training, we can get a preliminary caching strategy. To further improve caching performance, we propose a multiple-satellites collaboration based on deep reinforcement learning. This collaborative approach enhances the cache hit ratio and diminishes content request delay.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2866-2881"},"PeriodicalIF":4.7,"publicationDate":"2025-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144232005","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}

{"title":"DRL-Based Time-Varying Workload Scheduling With Priority and Resource Awareness","authors":"Qifeng Liu;Qilin Fan;Xu Zhang;Xiuhua Li;Kai Wang;Qingyu Xiong","doi":"10.1109/TNSM.2025.3559610","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3559610","url":null,"abstract":"With the proliferation of cloud services and the continuous growth in enterprises’ demand for dynamic multi-dimensional resources, the implementation of effective strategy for time-varying workload scheduling has become increasingly significant. In this paper, we propose a deep reinforcement learning (DRL)-based method for time-varying workload scheduling, aiming to allocate resources efficiently across servers in the cluster. Specifically, we integrate a classifier and queue scorer to construct a priority queue that exploits temporal resource utilization patterns across different workload classes. Then, we design parallel graph attention layers to capture the dimensional features and temporal dynamics of cloud server cluster. Moreover, we propose a DRL algorithm to generate scheduling strategies that can adapt to dynamic environments. Validation on real-world traces from Google cluster demonstrates that our method outperforms existing approaches in key metrics of cloud server cluster management.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2838-2852"},"PeriodicalIF":4.7,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144232148","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}

{"title":"ArchSentry: Enhanced Android Malware Detection via Hierarchical Semantic Extraction","authors":"Tianbo Wang;Mengyao Liu;Huacheng Li;Lei Zhao;Changnan Jiang;Chunhe Xia;Baojiang Cui","doi":"10.1109/TNSM.2025.3559255","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3559255","url":null,"abstract":"Android malware poses a significant challenge for mobile platforms. To evade detection, contemporary malware variants use API substitution or obfuscation techniques to hide malicious activities and mask their shallow semantic characteristics. However, existing research lacks analysis of the hierarchical semantic associated with Android apps. To address this problem, we propose ArchSentry, an enhanced Android malware detection via hierarchical semantic extraction. First, we select entities and their relationships relevant to Android software behavior through the software architecture and represent them using a heterogeneous graph. Then, we structure meta-paths to represent rich semantic information to achieve semantic enhancement and improve efficiency. Next, we design a meta-path semantic selection method based on KL Divergence to identify and eliminate redundant features. To achieve a comprehensive representation of the overall software semantics and improve performance, we construct a feature fusion approach based on Restricted Boltzmann Machines (RBM) and AutoEncoder (AE) during the pre-training phase, while preserving the probability distribution characteristics of various meta-paths. Finally, Deep Neural Networks (DNN) process fusion features for comprehensive feature sets. Experimental results on real-world application samples indicate that ArchSentry achieves a remarkable 99.2% detection rate for Android malware, with a low false positive rate below 1%. These results surpass the performance of current state-of-the-art approaches.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2822-2837"},"PeriodicalIF":4.7,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144229457","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}

{"title":"Joint Controller Placement and TDMA Scheduling in Software Defined Wireless Multihop Networks","authors":"Yiannis Papageorgiou;Merkouris Karaliopoulos;Kostas Choumas;Iordanis Koutsopoulos","doi":"10.1109/TNSM.2025.3559104","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3559104","url":null,"abstract":"We study TDMA-scheduled Software Defined Wireless Multihop Networks (SDWMNs), whereby the data traffic and SDN control messages share the same network links and TDMA resources. Since the topology of WMNs dynamically changes, maintaining a responsive SDN plane is essential for meeting data traffic rate requirements. Placing more SDN controllers reduces communication delays at the SDN layer and increases its responsiveness. However, it demands more TDMA resources and reduces the available ones for data traffic. We analyze this trade-off between data traffic performance and SDN layer responsiveness by delving into two distinct resource allocation mechanisms in the WMN, the SDN controller placement and TDMA scheduling. We capture their interaction into an optimization problem formulation, which aims at maximizing the SDN-responsiveness subject to data traffic rate requirements, topology conditions, and the available TDMA resources. We propose a novel heuristic for the hard-to-solve problem that leverages the network state information gathered at the SDN layer. We find that our heuristic can increase the SDN-responsiveness by 44% when varying the rate reserved for rate-elastic data traffic within 40% of what is nominally requested. The heuristic is modular in accommodating different controller placement algorithms and robust to different alternative for the SDN software implementation.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2807-2821"},"PeriodicalIF":4.7,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144232191","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}

{"title":"Bandwidth Efficient Cache Selection and Cache-Content Advertisement","authors":"Itamar Cohen","doi":"10.1109/TNSM.2025.3558835","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3558835","url":null,"abstract":"Caching is extensively used in various networking environments to optimize performance by reducing latency, bandwidth, and energy consumption. To optimize performance, caches often advertise their content using indicators, which are data structures that trade space efficiency for accuracy. However, this tradeoff introduces the risk of false indications. Existing solutions for cache content advertisement and cache selection often lead to inefficiencies, failing to adapt to dynamic network conditions. This paper introduces SALSA2, a Scalable Adaptive and Learning-based Selection and Advertisement Algorithm, which addresses these limitations through a dynamic and adaptive approach. SALSA2 accurately estimates mis-indication probabilities by considering inter-cache dependencies and dynamically adjusts the size and frequency of indicator advertisements to minimize transmission overhead while maintaining high accuracy. Our extensive simulation study, conducted using a variety of real-world cache traces, demonstrates that SALSA2 achieves up to 84% bandwidth savings compared to the state-of-the-art solution and close-to-optimal service cost in most scenarios. These results highlight SALSA2’s effectiveness in enhancing cache management, making it a robust and versatile solution for modern networking challenges.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2795-2806"},"PeriodicalIF":4.7,"publicationDate":"2025-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144232100","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}

{"title":"QoE-Aware Joint Visual and Haptic Signal Transmission With Adaptive Data Compression for Immersive Interactions in Human Digital Twin","authors":"Kun Wu;Jiayuan Chen;Lucheng Chen;Zili Liu;Changyan Yi;Shuai Xu;Junyi Wang;Jiawen Kang","doi":"10.1109/TNSM.2025.3558358","DOIUrl":"https://doi.org/10.1109/TNSM.2025.3558358","url":null,"abstract":"Human digital twin (HDT) is envisioned as a system interconnecting physical twins (PTs) in the real world with virtual twins (VTs) in the digital world, enabling advanced human-centric applications. Unlike optimizing the quality of experience (QoE) of users in single-modal signal transmission for conventional services, users’ QoE in multi-modal signal transmission required by HDT is difficult to guarantee. To tackle this, we study an optimization of QoE in multi-modal transmission, focusing on joint visual and haptic signal feedback transmissions from VT to its PT, for providing immersive interactions in HDT. To evaluate a synthesized performance of visual and haptic experiences, we design a comprehensive QoE model, taking into account video quality, continuous video quality switching rate and average haptic feedback error. Then, to maximize QoE with a guarantee on synchronization between visual and haptic signal transmissions, we dynamically optimize bandwidth allocation, bitrate and rendering mode of the video, and haptic signal’s compression threshold. To this end, we propose a deep reinforcement learning based algorithm, called VisHap. Furthermore, we build an HDT multi-modal interaction platform for collecting an authentic dataset, and by using it, we conduct experiments, showing that VisHap is not only feasible but also outperforms the counterparts.","PeriodicalId":13423,"journal":{"name":"IEEE Transactions on Network and Service Management","volume":"22 3","pages":"2780-2794"},"PeriodicalIF":4.7,"publicationDate":"2025-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144229555","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}