Ad Hoc Networks最新文献_第5页

A dual layer LSTM-CNN framework for real time and precise per-message intrusion detection in In-vehicle networks 基于LSTM-CNN的车载网络实时精确单消息入侵检测双层框架

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-12-04 DOI: 10.1016/j.adhoc.2025.104119

Yu Fu, Junhui She, Yinan Xu, Yihu Xu, Ziyi Wang, Yujing Wu

The rapid proliferation of Intelligent Connected Vehicles (ICVs) presents escalating cybersecurity challenges, particularly within Controller Area Networks (CAN), where traditional Intrusion Detection Systems (IDS) often fail to meet the stringent requirements for real-time and fine-grained anomaly detection under limited computational resources. This study introduces a novel lightweight dual-tier anomaly detection framework, termed Enhanced LSTM-CNN (ELC), which synergistically integrates temporal and spatial deep learning paradigms to address these limitations. The first tier employs an optimized hybrid architecture combining Long Short-Term Memory (LSTM) networks for temporal dependency modeling and Convolutional Neural Networks (CNNs) for spatial feature extraction, enabling rapid and accurate preliminary anomaly screening. The second tier utilizes an enhanced CNN classifier to perform refined multi-class identification of four prevalent attack types, including Denial of Service (DoS), Fuzzy, and RPM/GEAR spoofing, achieving an F1-score of 99.8 %. Comprehensive evaluations on real-world vehicular CAN datasets demonstrate that ELC attains an average per-message detection of 0.153 ms and sustains a processing throughput of 7000 messages per second, all within a power envelope of 7.3 W making it well suited for deployment in resource-constrained Electronic Control Units (ECUs). In addition, we validate ELC on the public 4TU CAN Bus Intrusion Dataset v2 and Survival Analysis Dataset maintaining comparable performance under cross-dataset settings and underscoring generalization and reproducibility. Unlike conventional batch-based approaches, ELC provides message-level granularity and sub-millisecond responsiveness, thereby ensuring timely threat mitigation within the 10 ms message interval constraints of CAN systems. These results indicate that the proposed framework holds strong potential as a practical and effective solution for real-time, embedded intrusion detection in resource-constrained vehicular environments.

智能网联汽车（icv）的快速发展带来了不断升级的网络安全挑战，特别是在控制器局域网（CAN）中，传统的入侵检测系统（IDS）通常无法满足在有限的计算资源下实时和细粒度异常检测的严格要求。本研究引入了一种新的轻量级双层异常检测框架，称为增强型LSTM-CNN (ELC)，它协同集成了时空深度学习范式来解决这些限制。第一层采用优化的混合架构，结合长短期记忆（LSTM）网络进行时间依赖性建模和卷积神经网络（cnn）进行空间特征提取，实现快速准确的初步异常筛选。第二层利用增强的CNN分类器对四种流行的攻击类型进行精细的多类识别，包括拒绝服务（DoS）、模糊和RPM/GEAR欺骗，达到了99.8%的f1分数。对实际车辆CAN数据集的综合评估表明，ELC实现了平均每消息0.153 ms的检测，并保持每秒7000条消息的处理吞吐量，所有这些都在7.3 W的功率范围内，使其非常适合部署在资源受限的电子控制单元（ecu）中。此外，我们在公共4TU CAN总线入侵数据集v2和生存分析数据集上验证了ELC，在跨数据集设置下保持了相当的性能，并强调了泛化和可重复性。与传统的基于批处理的方法不同，ELC提供消息级粒度和亚毫秒级响应，从而确保在CAN系统的10毫秒消息间隔限制内及时缓解威胁。这些结果表明，该框架具有强大的潜力，可作为资源受限车辆环境下实时嵌入式入侵检测的实用有效解决方案。

{"title":"A dual layer LSTM-CNN framework for real time and precise per-message intrusion detection in In-vehicle networks","authors":"Yu Fu, Junhui She, Yinan Xu, Yihu Xu, Ziyi Wang, Yujing Wu","doi":"10.1016/j.adhoc.2025.104119","DOIUrl":"10.1016/j.adhoc.2025.104119","url":null,"abstract":"<div><div>The rapid proliferation of Intelligent Connected Vehicles (ICVs) presents escalating cybersecurity challenges, particularly within Controller Area Networks (CAN), where traditional Intrusion Detection Systems (IDS) often fail to meet the stringent requirements for real-time and fine-grained anomaly detection under limited computational resources. This study introduces a novel lightweight dual-tier anomaly detection framework, termed Enhanced LSTM-CNN (ELC), which synergistically integrates temporal and spatial deep learning paradigms to address these limitations. The first tier employs an optimized hybrid architecture combining Long Short-Term Memory (LSTM) networks for temporal dependency modeling and Convolutional Neural Networks (CNNs) for spatial feature extraction, enabling rapid and accurate preliminary anomaly screening. The second tier utilizes an enhanced CNN classifier to perform refined multi-class identification of four prevalent attack types, including Denial of Service (DoS), Fuzzy, and RPM/GEAR spoofing, achieving an F1-score of 99.8 %. Comprehensive evaluations on real-world vehicular CAN datasets demonstrate that ELC attains an average per-message detection of 0.153 ms and sustains a processing throughput of 7000 messages per second, all within a power envelope of 7.3 W making it well suited for deployment in resource-constrained Electronic Control Units (ECUs). In addition, we validate ELC on the public 4TU CAN Bus Intrusion Dataset v2 and Survival Analysis Dataset maintaining comparable performance under cross-dataset settings and underscoring generalization and reproducibility. Unlike conventional batch-based approaches, ELC provides message-level granularity and sub-millisecond responsiveness, thereby ensuring timely threat mitigation within the 10 ms message interval constraints of CAN systems. These results indicate that the proposed framework holds strong potential as a practical and effective solution for real-time, embedded intrusion detection in resource-constrained vehicular environments.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"183 ","pages":"Article 104119"},"PeriodicalIF":4.8,"publicationDate":"2025-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738228","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

NL-MHP: Efficient and robust network localization algorithm in complex scenarios using maximum hop progress NL-MHP：基于最大跳数进展的复杂场景下高效鲁棒的网络定位算法

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-12-02 DOI: 10.1016/j.adhoc.2025.104110

Zhihao Dong , Xiaoyong Yan , Jian Zhou

Accurate node localization is critical for wireless network applications, yet existing algorithms struggle with low communication efficiency, inaccurate distance measurement, and unstable localization in irregular multi-hop networks. To address these challenges, this paper proposes an Efficient and Robust Network Localization algorithm using Maximum Hop Progress (NL-MHP). NL-MHP integrates routing exploration and distance estimation into a single phase, significantly reducing communication overhead. It further introduces an adaptive hop count threshold to filter out erroneous distances and constructs constrained subregions for each unlocated node to ensure localization stability. The Coati optimization algorithm with Sobol sequence is then employed within these subregions to determine the optimal node location. Comprehensive simulations demonstrate that NL-MHP outperforms existing methods in efficiency, accuracy, and stability for irregular networks. Quantitatively, it improves distance estimation accuracy by 26.7% to 34.8% and localization accuracy by 31.4% to 51.7% compared to three state-of-the-art algorithms.

准确的节点定位是无线网络应用的关键，但在不规则多跳网络中，现有算法存在通信效率低、距离测量不准确、定位不稳定等问题。为了解决这些问题，本文提出了一种基于最大跳数进展（NL-MHP）的高效鲁棒网络定位算法。NL-MHP将路由探索和距离估计集成到一个阶段，显著降低了通信开销。它进一步引入自适应跳数阈值来过滤错误距离，并为每个未定位节点构建约束子区域以确保定位稳定性。然后在这些子区域内使用Sobol序列的Coati优化算法来确定最优节点位置。综合仿真表明，对于不规则网络，NL-MHP在效率、精度和稳定性方面都优于现有方法。在数量上，与三种最先进的算法相比，它将距离估计精度提高了26.7%至34.8%，定位精度提高了31.4%至51.7%。

{"title":"NL-MHP: Efficient and robust network localization algorithm in complex scenarios using maximum hop progress","authors":"Zhihao Dong , Xiaoyong Yan , Jian Zhou","doi":"10.1016/j.adhoc.2025.104110","DOIUrl":"10.1016/j.adhoc.2025.104110","url":null,"abstract":"<div><div>Accurate node localization is critical for wireless network applications, yet existing algorithms struggle with low communication efficiency, inaccurate distance measurement, and unstable localization in irregular multi-hop networks. To address these challenges, this paper proposes an Efficient and Robust <u>N</u>etwork <u>L</u>ocalization algorithm using <u>M</u>aximum <u>H</u>op <u>P</u>rogress (NL-MHP). NL-MHP integrates routing exploration and distance estimation into a single phase, significantly reducing communication overhead. It further introduces an adaptive hop count threshold to filter out erroneous distances and constructs constrained subregions for each unlocated node to ensure localization stability. The Coati optimization algorithm with Sobol sequence is then employed within these subregions to determine the optimal node location. Comprehensive simulations demonstrate that NL-MHP outperforms existing methods in efficiency, accuracy, and stability for irregular networks. Quantitatively, it improves distance estimation accuracy by 26.7% to 34.8% and localization accuracy by 31.4% to 51.7% compared to three state-of-the-art algorithms.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"183 ","pages":"Article 104110"},"PeriodicalIF":4.8,"publicationDate":"2025-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145694343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Dependent tasks joint scheduling and offloading for edge computing based on deep reinforcement learning 基于深度强化学习的边缘计算相关任务联合调度与卸载

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-12-02 DOI: 10.1016/j.adhoc.2025.104111

Debin Wei , Jinglong Wen , Pingduo Xu , Huaifeng Shi , Chengsheng Pan

In the mobile edge computing (MEC) scenario, numerous complex applications consist of dependent tasks. Efficient offloading of these applications is essential for reducing latency and minimizing terminal energy consumption. However, existing studies typically employ a decoupled decision-making paradigm, where the task scheduling sequence is predetermined before independently determining the offloading location. This approach separates the scheduling and offloading processes, significantly limiting the exploration of the strategy space and hindering the identification of the global optimal solution. To address these limitations, we propose a joint scheduling and offloading algorithm based on Proximal Policy Optimization (JSO-PPO). We construct an integrated Markov Decision Process (MDP) model and introduce an action masking mechanism, unifying task scheduling and location offloading into a single end-to-end decision. Furthermore, to enhance the algorithm’s performance and stability, the JSO-PPO integrates Deep Dense Architectures in Reinforcement Learning (D2RL) for superior state representation and introduces an adaptive penalty term into its objective function for more stable convergence. Simulation results demonstrate that, compared to multiple existing algorithms, the proposed JSO-PPO achieves significant improvements in minimizing the weighted sum of application finish latency and user equipment energy consumption. These findings validate the efficiency and robustness of our joint optimization paradigm in dynamic and complex edge environments.

在移动边缘计算（MEC）场景中，许多复杂的应用程序由相关任务组成。有效地卸载这些应用程序对于减少延迟和最小化终端能耗至关重要。然而，现有研究通常采用解耦决策范式，即在独立确定卸载位置之前预先确定任务调度顺序。这种方法分离了调度和卸载过程，严重限制了策略空间的探索，阻碍了全局最优解的识别。为了解决这些限制，我们提出了一种基于近端策略优化（JSO-PPO）的联合调度和卸载算法。我们构建了一个集成的马尔可夫决策过程（MDP）模型，并引入了动作掩蔽机制，将任务调度和位置卸载统一到一个端到端决策中。此外，为了提高算法的性能和稳定性，JSO-PPO在强化学习（D2RL）中集成了深度密集架构，以获得更好的状态表示，并在其目标函数中引入了自适应惩罚项，以实现更稳定的收敛。仿真结果表明，与现有的多种算法相比，所提出的JSO-PPO算法在最小化应用完成延迟加权和用户设备能耗方面取得了显著的进步。这些结果验证了我们的联合优化范式在动态和复杂边缘环境中的有效性和鲁棒性。

{"title":"Dependent tasks joint scheduling and offloading for edge computing based on deep reinforcement learning","authors":"Debin Wei , Jinglong Wen , Pingduo Xu , Huaifeng Shi , Chengsheng Pan","doi":"10.1016/j.adhoc.2025.104111","DOIUrl":"10.1016/j.adhoc.2025.104111","url":null,"abstract":"<div><div>In the mobile edge computing (MEC) scenario, numerous complex applications consist of dependent tasks. Efficient offloading of these applications is essential for reducing latency and minimizing terminal energy consumption. However, existing studies typically employ a decoupled decision-making paradigm, where the task scheduling sequence is predetermined before independently determining the offloading location. This approach separates the scheduling and offloading processes, significantly limiting the exploration of the strategy space and hindering the identification of the global optimal solution. To address these limitations, we propose a joint scheduling and offloading algorithm based on Proximal Policy Optimization (JSO-PPO). We construct an integrated Markov Decision Process (MDP) model and introduce an action masking mechanism, unifying task scheduling and location offloading into a single end-to-end decision. Furthermore, to enhance the algorithm’s performance and stability, the JSO-PPO integrates Deep Dense Architectures in Reinforcement Learning (D2RL) for superior state representation and introduces an adaptive penalty term into its objective function for more stable convergence. Simulation results demonstrate that, compared to multiple existing algorithms, the proposed JSO-PPO achieves significant improvements in minimizing the weighted sum of application finish latency and user equipment energy consumption. These findings validate the efficiency and robustness of our joint optimization paradigm in dynamic and complex edge environments.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"183 ","pages":"Article 104111"},"PeriodicalIF":4.8,"publicationDate":"2025-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738225","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

RAG-HIDS: A multi-relational graph-based hierarchical intrusion detection system for in-vehicle networks 基于多关系图的车载网络分层入侵检测系统

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-11-29 DOI: 10.1016/j.adhoc.2025.104108

Hai Lin, Xi Yu, Zhihong Chen, Yue Cao

With the increasing complexity and connectivity of modern vehicles, the security of in-vehicle networks has become a critical concern. The Controller Area Network (CAN), as the primary communication protocol, lacks authentication and encryption, making it highly vulnerable to cyberattacks. Intrusion detection systems (IDSs) serve as a vital line of defense by monitoring CAN traffic and identifying suspicious behavior. While graph-based IDSs can model complex relationships among CAN traffic, most existing graph-based methods operate at coarse detection levels such as window or CAN ID level. This limits their ability to detect fine-grained or stealthy attacks. To overcome these limitations, we propose RAG-HIDS, a novel Relation-Aware Graph-based Hierarchical Intrusion Detection System for CAN bus traffic. Our method constructs a multi-relational graph that captures both short-term temporal dependencies and long-term functional consistency among messages. Based on this representation, we design a hierarchical detection framework that first performs rapid intrusion screening at the window level, followed by fine-grained anomaly localization at the message level. Experiments on Car-Hacking dataset show that RAG-HIDS achieves 100% accuracy at the window level and over 99.6% at the message level across multiple attack types. And we demonstrate the model’s robustness across different vehicle types through multi-vehicle training and per-vehicle evaluation on the Survival dataset.

随着现代车辆的日益复杂和互联，车载网络的安全性已成为一个重要问题。控制器区域网络（CAN）作为主要的通信协议，缺乏认证和加密，极易受到网络攻击。入侵检测系统（ids）通过监控CAN流量和识别可疑行为作为重要的防线。虽然基于图的ids可以对can流量之间的复杂关系进行建模，但大多数现有的基于图的方法都是在粗糙的检测级别（如窗口或can ID级别）上运行的。这限制了它们检测细粒度或隐蔽攻击的能力。为了克服这些限制，我们提出了一种基于关系感知图的CAN总线流量分层入侵检测系统RAG-HIDS。我们的方法构建了一个多关系图，可以捕获消息之间的短期时间依赖性和长期功能一致性。基于这种表示，我们设计了一个分层检测框架，首先在窗口级执行快速入侵筛选，然后在消息级执行细粒度异常定位。在汽车黑客数据集上的实验表明，在多种攻击类型下，rags - hids在窗口级别达到100%的准确率，在消息级别达到99.6%以上。我们通过多车训练和对生存数据集的每车评估来证明模型在不同车辆类型上的鲁棒性。

{"title":"RAG-HIDS: A multi-relational graph-based hierarchical intrusion detection system for in-vehicle networks","authors":"Hai Lin, Xi Yu, Zhihong Chen, Yue Cao","doi":"10.1016/j.adhoc.2025.104108","DOIUrl":"10.1016/j.adhoc.2025.104108","url":null,"abstract":"<div><div>With the increasing complexity and connectivity of modern vehicles, the security of in-vehicle networks has become a critical concern. The Controller Area Network (CAN), as the primary communication protocol, lacks authentication and encryption, making it highly vulnerable to cyberattacks. Intrusion detection systems (IDSs) serve as a vital line of defense by monitoring CAN traffic and identifying suspicious behavior. While graph-based IDSs can model complex relationships among CAN traffic, most existing graph-based methods operate at coarse detection levels such as window or CAN ID level. This limits their ability to detect fine-grained or stealthy attacks. To overcome these limitations, we propose RAG-HIDS, a novel Relation-Aware Graph-based Hierarchical Intrusion Detection System for CAN bus traffic. Our method constructs a multi-relational graph that captures both short-term temporal dependencies and long-term functional consistency among messages. Based on this representation, we design a hierarchical detection framework that first performs rapid intrusion screening at the window level, followed by fine-grained anomaly localization at the message level. Experiments on Car-Hacking dataset show that RAG-HIDS achieves 100% accuracy at the window level and over 99.6% at the message level across multiple attack types. And we demonstrate the model’s robustness across different vehicle types through multi-vehicle training and per-vehicle evaluation on the Survival dataset.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"183 ","pages":"Article 104108"},"PeriodicalIF":4.8,"publicationDate":"2025-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145651806","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

PGN-MO-DDQN: A preference-driven multi-objective offloading algorithm for mobile edge video analytics PGN-MO-DDQN：一种偏好驱动的移动边缘视频分析多目标卸载算法

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-11-28 DOI: 10.1016/j.adhoc.2025.104100

Honghai Wu, Chenyang Wang, Ling Xing, Huahong Ma, Ruijuan Zheng, Xiaoli Song

With the rapid proliferation of intelligent video analytics in real-time applications such as the Internet of Vehicles and urban surveillance, Mobile Edge Computing (MEC) systems face persistent challenges, including complex inter-frame dependencies, severe bandwidth fluctuations, and intensive resource contention. Conventional offloading strategies based on fixed rules or static weights are often inadequate for maintaining robustness in scenarios with conflicting multi-objective requirements. To address these challenges, this paper proposes PGN-MO-DDQN, an adaptive multi-objective reinforcement learning framework for task offloading and scheduling, driven by a Preference Generation Network (PGN). The proposed framework leverages the multi-objective architecture of Double-DQN and incorporates a PGN to automatically produce dynamic weight vectors for latency, load balancing, and analytical accuracy (mean average precision, mAP) according to real-time system states, while dynamically determining the offloading ratio between local and edge processing to balance communication latency and computational load, thereby enabling adaptive multi-objective optimization. Experimental evaluations conducted on the EdgeSimPy platform show that, compared with representative baseline algorithms such as Greedy, Fixed-DDQN, HRL-V2I, and TOLB, PGN-MO-DDQN reduces average task latency by 17%, decreases server load imbalance by 13%, and improves mAP by 5.8%. Overall, the proposed framework provides a robust solution for adaptive multi-objective video offloading, effectively balancing latency, accuracy, and resource use in dynamic edge environments.

随着智能视频分析在实时应用（如车联网和城市监控）中的迅速普及，移动边缘计算（MEC）系统面临着持续的挑战，包括复杂的帧间依赖关系、严重的带宽波动和密集的资源争用。传统的基于固定规则或静态权重的卸载策略通常不足以在具有相互冲突的多目标需求的场景中保持鲁棒性。为了解决这些挑战，本文提出了PGN- mo - ddqn，这是一个由偏好生成网络（PGN）驱动的用于任务卸载和调度的自适应多目标强化学习框架。该框架利用Double-DQN的多目标架构，结合PGN，根据实时系统状态自动生成时延、负载均衡和分析精度（mean average precision, mAP）的动态权重向量，同时动态确定本地和边缘处理之间的卸载比例，平衡通信时延和计算负载，从而实现自适应多目标优化。在EdgeSimPy平台上进行的实验评估表明，与Greedy、Fixed-DDQN、HRL-V2I、TOLB等代表性基线算法相比，PGN-MO-DDQN平均任务延迟降低17%，服务器负载不平衡降低13%，mAP提高5.8%。总体而言，所提出的框架为自适应多目标视频卸载提供了一个强大的解决方案，有效地平衡了动态边缘环境中的延迟、准确性和资源使用。

{"title":"PGN-MO-DDQN: A preference-driven multi-objective offloading algorithm for mobile edge video analytics","authors":"Honghai Wu, Chenyang Wang, Ling Xing, Huahong Ma, Ruijuan Zheng, Xiaoli Song","doi":"10.1016/j.adhoc.2025.104100","DOIUrl":"10.1016/j.adhoc.2025.104100","url":null,"abstract":"<div><div>With the rapid proliferation of intelligent video analytics in real-time applications such as the Internet of Vehicles and urban surveillance, Mobile Edge Computing (MEC) systems face persistent challenges, including complex inter-frame dependencies, severe bandwidth fluctuations, and intensive resource contention. Conventional offloading strategies based on fixed rules or static weights are often inadequate for maintaining robustness in scenarios with conflicting multi-objective requirements. To address these challenges, this paper proposes PGN-MO-DDQN, an adaptive multi-objective reinforcement learning framework for task offloading and scheduling, driven by a Preference Generation Network (PGN). The proposed framework leverages the multi-objective architecture of Double-DQN and incorporates a PGN to automatically produce dynamic weight vectors for latency, load balancing, and analytical accuracy (mean average precision, mAP) according to real-time system states, while dynamically determining the offloading ratio between local and edge processing to balance communication latency and computational load, thereby enabling adaptive multi-objective optimization. Experimental evaluations conducted on the EdgeSimPy platform show that, compared with representative baseline algorithms such as Greedy, Fixed-DDQN, HRL-V2I, and TOLB, PGN-MO-DDQN reduces average task latency by 17%, decreases server load imbalance by 13%, and improves mAP by 5.8%. Overall, the proposed framework provides a robust solution for adaptive multi-objective video offloading, effectively balancing latency, accuracy, and resource use in dynamic edge environments.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"182 ","pages":"Article 104100"},"PeriodicalIF":4.8,"publicationDate":"2025-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145693912","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Protecting autonomous systems from GPS spoofing with a machine learning-driven approach 利用机器学习驱动的方法保护自主系统免受GPS欺骗

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-11-26 DOI: 10.1016/j.adhoc.2025.104101

Arslan Shafique , Abid Mehmood , Moatsum Alawida , Shehzad Ashraf Chaudhry

With the rapid evolution of interactive multimedia systems, ensuring strong security measures has become increasingly vital. Autonomous platforms, such as drones, are vulnerable to sophisticated cyber threats, including jamming and spoofing attacks. One common spoofing strategy involves manipulating Global Positioning System (GPS) signals. By broadcasting counterfeit signals, attackers can deceive drone navigation systems. To mitigate these risks, this research introduces a machine learning-based framework aimed at intelligently detecting spoofing attempts. The approach employs signal characteristics that reflect variations in jitter, shimmer, and frequency modulations, rooted in mathematical analysis. A private dataset is used for this work. This dataset, developed by our research team, is collected under varied environmental conditions, such as during daylight and in low-light settings, over multiple sessions. It categorizes signal statistics into three distinct ranges: the initial and final segments suggest spoofed signals, whereas the middle range corresponds to genuine ones. Further, using signal reception strength (SRS) values, additional data was sourced from trusted Long Range Wide Area Network (LoRaWAN) devices. This information supports the development of a singular-class support vector machine (S-CSVM) classifier for spoofing detection. For training purposes, the dataset was partitioned into two subsets: a training set (

T_{t r a i n}

) and a testing set (

T_{t e s t}

). The performance of the model is evaluated using standard metrics, including precision, recall, F-score, and overall accuracy. By utilizing all available features, the model achieves its highest scores: 99.99% precision, 99.77% recall, and 99.95% F-score. The highest accuracy of 99.22% is achieved when all features are selected, and the distance between LoRaWAN devices and the monitoring device ranges from 6 to 8 m, outperforming the results obtained through feature selection in the ablation study. A thorough evaluation further highlights how the proposed ML-based solution outperforms existing methods.

随着交互式多媒体系统的快速发展，确保强大的安全措施变得越来越重要。无人机等自主平台容易受到复杂的网络威胁，包括干扰和欺骗攻击。一种常见的欺骗策略涉及操纵全球定位系统（GPS）信号。通过传播伪造信号，攻击者可以欺骗无人机导航系统。为了降低这些风险，本研究引入了一种基于机器学习的框架，旨在智能地检测欺骗企图。该方法采用反映抖动、闪烁和频率调制变化的信号特性，植根于数学分析。这项工作使用了一个私有数据集。这个数据集是由我们的研究团队开发的，是在不同的环境条件下收集的，比如在白天和低光环境下，在多个会话中。它将信号统计分为三个不同的范围：初始和最终段表示欺骗信号，而中间范围对应于真实信号。此外，使用信号接收强度（SRS）值，从可信的远程广域网（LoRaWAN）设备获取额外数据。这些信息支持开发用于欺骗检测的奇异类支持向量机（S-CSVM）分类器。为了训练目的，数据集被划分为两个子集：训练集（Ttrain）和测试集（Ttest）。使用标准指标评估模型的性能，包括精度、召回率、f分数和总体准确性。通过利用所有可用的特征，该模型达到了最高的分数：99.99%的精度，99.77%的召回率和99.95%的f分数。当选择所有特征时，LoRaWAN设备与监测设备之间的距离为6 ~ 8 m，准确率最高，达到99.22%，优于烧蚀研究中通过特征选择获得的结果。全面的评估进一步强调了所提出的基于ml的解决方案如何优于现有方法。

{"title":"Protecting autonomous systems from GPS spoofing with a machine learning-driven approach","authors":"Arslan Shafique , Abid Mehmood , Moatsum Alawida , Shehzad Ashraf Chaudhry","doi":"10.1016/j.adhoc.2025.104101","DOIUrl":"10.1016/j.adhoc.2025.104101","url":null,"abstract":"<div><div>With the rapid evolution of interactive multimedia systems, ensuring strong security measures has become increasingly vital. Autonomous platforms, such as drones, are vulnerable to sophisticated cyber threats, including jamming and spoofing attacks. One common spoofing strategy involves manipulating Global Positioning System (GPS) signals. By broadcasting counterfeit signals, attackers can deceive drone navigation systems. To mitigate these risks, this research introduces a machine learning-based framework aimed at intelligently detecting spoofing attempts. The approach employs signal characteristics that reflect variations in jitter, shimmer, and frequency modulations, rooted in mathematical analysis. A private dataset is used for this work. This dataset, developed by our research team, is collected under varied environmental conditions, such as during daylight and in low-light settings, over multiple sessions. It categorizes signal statistics into three distinct ranges: the initial and final segments suggest spoofed signals, whereas the middle range corresponds to genuine ones. Further, using signal reception strength (SRS) values, additional data was sourced from trusted Long Range Wide Area Network (LoRaWAN) devices. This information supports the development of a singular-class support vector machine (S-CSVM) classifier for spoofing detection. For training purposes, the dataset was partitioned into two subsets: a training set (<span><math><msub><mrow><mi>T</mi></mrow><mrow><mi>t</mi><mi>r</mi><mi>a</mi><mi>i</mi><mi>n</mi></mrow></msub></math></span>) and a testing set (<span><math><msub><mrow><mi>T</mi></mrow><mrow><mi>t</mi><mi>e</mi><mi>s</mi><mi>t</mi></mrow></msub></math></span>). The performance of the model is evaluated using standard metrics, including precision, recall, F-score, and overall accuracy. By utilizing all available features, the model achieves its highest scores: 99.99% precision, 99.77% recall, and 99.95% F-score. The highest accuracy of 99.22% is achieved when all features are selected, and the distance between LoRaWAN devices and the monitoring device ranges from 6 to 8 m, outperforming the results obtained through feature selection in the ablation study. A thorough evaluation further highlights how the proposed ML-based solution outperforms existing methods.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"183 ","pages":"Article 104101"},"PeriodicalIF":4.8,"publicationDate":"2025-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145651805","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Dynamic deployment of UAVs for temporary networks using multi-criteria decision-making 基于多准则决策的临时网络无人机动态部署

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-11-26 DOI: 10.1016/j.adhoc.2025.104096

Flávio Henry Ferreira , Fabrício J.B. Barros , Miércio C.A. Neto , Arun Narayanan , Pedro H.J. Nardelli , Jasmine P.L. Araújo

Unmanned Aerial Vehicle Base Stations (UAV-BSs) are effective to support mobile wireless systems in situations where an unusually high density of users require enhanced coverage and capacity such as in large sport events or music festivals. However, the joint deployment problem of UAV-BSs is NP-hard, and the optimization methods used to solve such a class of problems are often too slow for (quasi-)real-time applications when the density of users and the number of UAV-BS are high. This inefficiency creates a need for more adaptable methods to solve the UAV-BS positioning problem. This paper proposes a solution by transforming the UAV-BSs’ placement problem into a decision-making process using the Analytic Hierarchy Process (AHP) method. Our solution considers that all UAV-BSs move along scanning points with a predetermined path, each with a minimal distance from one another. The proposed algorithm, named UAV-AHP, efficiently determines the UAV-BSs’ positions, thereby improving the network performance based on (quasi-) real-time acquisition of user signals at scanning points. For the high-density scenarios under investigation, our numerical results demonstrate that UAV-AHP outperforms commonly used heuristics that sub-optimally solve NP-hard problems, namely, cuckoo search (CS), particle swarm optimization (PSO), and a genetic algorithm (NSGA-II). The proposed method for UAV-BS deployment requires considerably lower running times to find satisfactory solutions than CS, PSO, and NSGA-II.

无人机基站（UAV-BSs）在用户密度异常高、需要增强覆盖和容量的情况下（如大型体育赛事或音乐节）可有效支持移动无线系统。然而，无人机- bs的联合部署问题是np困难的，当用户密度和无人机- bs数量都很高时，用于解决这类问题的优化方法对于（准）实时应用来说往往太慢。这种低效率导致需要更具适应性的方法来解决无人机- bs定位问题。本文利用层次分析法（AHP）将无人机- bss的布局问题转化为决策问题，提出了解决方案。我们的解决方案考虑所有无人机- bss沿着预定路径的扫描点移动，每个扫描点之间的距离最小。所提出的算法称为UAV-AHP，该算法有效地确定了UAV-BSs的位置，从而提高了基于（准）实时获取扫描点用户信号的网络性能。对于研究中的高密度场景，我们的数值结果表明，UAV-AHP优于常用的启发式算法，即布谷鸟搜索（CS），粒子群优化（PSO）和遗传算法（NSGA-II）。与CS、PSO和NSGA-II相比，所提出的无人机- bs部署方法需要相当低的运行时间来找到令人满意的解决方案。

{"title":"Dynamic deployment of UAVs for temporary networks using multi-criteria decision-making","authors":"Flávio Henry Ferreira , Fabrício J.B. Barros , Miércio C.A. Neto , Arun Narayanan , Pedro H.J. Nardelli , Jasmine P.L. Araújo","doi":"10.1016/j.adhoc.2025.104096","DOIUrl":"10.1016/j.adhoc.2025.104096","url":null,"abstract":"<div><div>Unmanned Aerial Vehicle Base Stations (UAV-BSs) are effective to support mobile wireless systems in situations where an unusually high density of users require enhanced coverage and capacity such as in large sport events or music festivals. However, the joint deployment problem of UAV-BSs is NP-hard, and the optimization methods used to solve such a class of problems are often too slow for (quasi-)real-time applications when the density of users and the number of UAV-BS are high. This inefficiency creates a need for more adaptable methods to solve the UAV-BS positioning problem. This paper proposes a solution by transforming the UAV-BSs’ placement problem into a decision-making process using the Analytic Hierarchy Process (AHP) method. Our solution considers that all UAV-BSs move along scanning points with a predetermined path, each with a minimal distance from one another. The proposed algorithm, named UAV-AHP, efficiently determines the UAV-BSs’ positions, thereby improving the network performance based on (quasi-) real-time acquisition of user signals at scanning points. For the high-density scenarios under investigation, our numerical results demonstrate that UAV-AHP outperforms commonly used heuristics that sub-optimally solve NP-hard problems, namely, cuckoo search (CS), particle swarm optimization (PSO), and a genetic algorithm (NSGA-II). The proposed method for UAV-BS deployment requires considerably lower running times to find satisfactory solutions than CS, PSO, and NSGA-II.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"182 ","pages":"Article 104096"},"PeriodicalIF":4.8,"publicationDate":"2025-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145624178","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Co-Task: Collaborative Task Execution for UAV-as-a-Service using Stackelberg game 协同任务：使用Stackelberg游戏的无人机即服务协同任务执行

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-11-25 DOI: 10.1016/j.adhoc.2025.104099

Prince Kumar, Arijit Roy

This work proposes a Collaborative Task Execution (Co-Task) scheme for the UAV-as-a-Service (UaaS) platform. The UaaS platform enables multiple Unmanned Aerial Vehicles (UAVs) to collaboratively serve diverse Internet of Things (IoT) applications, thereby enhancing resource utilization, reducing energy consumption, and improving overall task efficiency and system reliability. However, effective collaboration among UAVs remains challenging due to task transfer delays, execution costs, and the presence of multiple UAV owners with competing economic interests. The problem becomes more challenging in the presence of multiple UAV owners and their competition to participate in the platform to earn monetary benefits. To address these challenges, Co-Task scheme ensures collaborative task execution among UAVs, even when they belong to different owners. We apply Stackelberg’s game-theoretic approach to design collaborative task execution in the UaaS platform, where a UAV with limited computational capacity (the leader) offloads a portion of its task to another UAV (the follower) with available resources. The leader and follower jointly optimize resource allocation and pricing to achieve efficient task execution and fair incentive distribution. Using gradient descent optimization and Nash equilibrium, we derive the optimal resource allocation and pricing strategies that ensure stable and efficient collaboration, where no UAV has an incentive to deviate unilaterally. To evaluate the effectiveness of Co-Task, we conducted extensive simulations and compared its performance with state-of-the-art approaches, – Gradient-based Iterative Search Algorithm (GISA), Game Theory Pricing Strategy (GTPS), and Heuristic with Learning Placement (HLP). The results demonstrate that Co-Task improves the leader’s utility by 7.51%.

本工作提出了一种用于无人机即服务（UaaS）平台的协同任务执行（Co-Task）方案。UaaS平台使多架无人机（uav）能够协同服务于各种物联网（IoT）应用，从而提高资源利用率，降低能耗，提高整体任务效率和系统可靠性。然而，由于任务转移延迟、执行成本和多个具有竞争经济利益的无人机所有者的存在，无人机之间的有效协作仍然具有挑战性。在多个无人机所有者和他们竞争参与平台以赚取金钱利益的情况下，这个问题变得更具挑战性。为了应对这些挑战，协同任务方案确保了无人机之间的协同任务执行，即使它们属于不同的所有者。我们应用Stackelberg的博弈论方法来设计UaaS平台中的协同任务执行，其中计算能力有限的无人机（领导者）将其任务的一部分卸载给具有可用资源的另一架无人机（追随者）。领导者和追随者共同优化资源配置和定价，以实现高效的任务执行和公平的激励分配。利用梯度下降优化和纳什均衡，我们推导了保证稳定高效协作的最优资源分配和定价策略，其中没有无人机有单方面偏离的动机。为了评估Co-Task的有效性，我们进行了广泛的模拟，并将其性能与最先进的方法进行了比较，这些方法包括基于梯度的迭代搜索算法（GISA）、博弈论定价策略（GTPS）和启发式学习放置（HLP）。结果表明，协同任务使领导者的效用提高了7.51%。

{"title":"Co-Task: Collaborative Task Execution for UAV-as-a-Service using Stackelberg game","authors":"Prince Kumar, Arijit Roy","doi":"10.1016/j.adhoc.2025.104099","DOIUrl":"10.1016/j.adhoc.2025.104099","url":null,"abstract":"<div><div>This work proposes a Collaborative Task Execution (Co-Task) scheme for the UAV-as-a-Service (UaaS) platform. The UaaS platform enables multiple Unmanned Aerial Vehicles (UAVs) to collaboratively serve diverse Internet of Things (IoT) applications, thereby enhancing resource utilization, reducing energy consumption, and improving overall task efficiency and system reliability. However, effective collaboration among UAVs remains challenging due to task transfer delays, execution costs, and the presence of multiple UAV owners with competing economic interests. The problem becomes more challenging in the presence of multiple UAV owners and their competition to participate in the platform to earn monetary benefits. To address these challenges, Co-Task scheme ensures collaborative task execution among UAVs, even when they belong to different owners. We apply Stackelberg’s game-theoretic approach to design collaborative task execution in the UaaS platform, where a UAV with limited computational capacity (the leader) offloads a portion of its task to another UAV (the follower) with available resources. The leader and follower jointly optimize resource allocation and pricing to achieve efficient task execution and fair incentive distribution. Using gradient descent optimization and Nash equilibrium, we derive the optimal resource allocation and pricing strategies that ensure stable and efficient collaboration, where no UAV has an incentive to deviate unilaterally. To evaluate the effectiveness of Co-Task, we conducted extensive simulations and compared its performance with state-of-the-art approaches, – Gradient-based Iterative Search Algorithm (GISA), Game Theory Pricing Strategy (GTPS), and Heuristic with Learning Placement (HLP). The results demonstrate that Co-Task improves the leader’s utility by 7.51%.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"182 ","pages":"Article 104099"},"PeriodicalIF":4.8,"publicationDate":"2025-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145624179","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Resource allocation and trajectory optimization for air-ground MEC systems with dual connectivity in unlicensed spectrum 无许可频谱双连接地空MEC系统的资源分配与轨迹优化

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-11-20 DOI: 10.1016/j.adhoc.2025.104098

Errong Pei, Niexin Xiang, Chenkai Ren

Unmanned Aerial Vehicles (UAVs) have emerged as pivotal components in Air-Ground Mobile Edge Computing (AGMEC) systems, leveraging their line-of-sight connectivity and maneuverability to enhance service coverage. In order to mitigate the growing spectrum scarcity, AGMEC systems are increasingly adopting unlicensed bands, though this raises interference concerns for incumbent users such as WiFi systems. To address these issues, this paper proposes three key innovations: (1) An interference-aware hybrid fairness spectrum sharing mechanism that ensures proportional fairness between AGMEC and WiFi systems while maintaining quality-of-service thresholds. (2) A dual-connectivity-enhanced data offloading scheme that integrates dual connectivity (DC) with frequency-division multiple access (FDMA), enabling terrestrial mobile users (MUs) to dynamically utilize both UAV and ground base station links for load-adaptive resource splitting. (3) A joint optimization framework that maximizes the dual-link weighted average rate through co-design of UAV trajectory, power allocation, and subchannel bandwidth assignment. The resulting mixed-integer non-convex problem is systematically decomposed via block coordinate descent (BCD) and solved iteratively using successive convex approximation (SCA). Simulation results show that our approach achieves a superior weighted sum rate compared to benchmark schemes, offering an effective solution for efficient AGMEC operation in shared unlicensed environments.

无人机（uav）已经成为空地移动边缘计算（AGMEC）系统中的关键组件，利用其视距连接和机动性来增强服务覆盖范围。为了缓解日益增长的频谱短缺，AGMEC系统越来越多地采用未经许可的频段，尽管这引起了对现有用户（如WiFi系统）的干扰担忧。为了解决这些问题，本文提出了三个关键创新：(1)干扰感知混合公平频谱共享机制，确保AGMEC和WiFi系统之间的比例公平，同时保持服务质量阈值。(2)一种双连接增强数据分流方案，该方案将双连接（DC）与频分多址（FDMA）相结合，使地面移动用户（mu）能够动态利用无人机和地面基站链路进行负载自适应资源分流。(3)通过无人机轨迹、功率分配和子信道带宽分配协同设计，实现双链路加权平均速率最大化的联合优化框架。采用分块坐标下降法（BCD）对混合整数非凸问题进行系统分解，并采用逐次凸逼近法（SCA）进行迭代求解。仿真结果表明，与基准方案相比，我们的方法获得了更高的加权和速率，为共享无许可环境下的高效AGMEC操作提供了有效的解决方案。

{"title":"Resource allocation and trajectory optimization for air-ground MEC systems with dual connectivity in unlicensed spectrum","authors":"Errong Pei, Niexin Xiang, Chenkai Ren","doi":"10.1016/j.adhoc.2025.104098","DOIUrl":"10.1016/j.adhoc.2025.104098","url":null,"abstract":"<div><div>Unmanned Aerial Vehicles (UAVs) have emerged as pivotal components in Air-Ground Mobile Edge Computing (AGMEC) systems, leveraging their line-of-sight connectivity and maneuverability to enhance service coverage. In order to mitigate the growing spectrum scarcity, AGMEC systems are increasingly adopting unlicensed bands, though this raises interference concerns for incumbent users such as WiFi systems. To address these issues, this paper proposes three key innovations: (1) An interference-aware hybrid fairness spectrum sharing mechanism that ensures proportional fairness between AGMEC and WiFi systems while maintaining quality-of-service thresholds. (2) A dual-connectivity-enhanced data offloading scheme that integrates dual connectivity (DC) with frequency-division multiple access (FDMA), enabling terrestrial mobile users (MUs) to dynamically utilize both UAV and ground base station links for load-adaptive resource splitting. (3) A joint optimization framework that maximizes the dual-link weighted average rate through co-design of UAV trajectory, power allocation, and subchannel bandwidth assignment. The resulting mixed-integer non-convex problem is systematically decomposed via block coordinate descent (BCD) and solved iteratively using successive convex approximation (SCA). Simulation results show that our approach achieves a superior weighted sum rate compared to benchmark schemes, offering an effective solution for efficient AGMEC operation in shared unlicensed environments.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"182 ","pages":"Article 104098"},"PeriodicalIF":4.8,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145624181","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Efficient joint optimization of task offloading and resource allocation for LEO satellite edge computing 低轨道卫星边缘计算任务卸载与资源分配的高效联合优化

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2025-11-19 DOI: 10.1016/j.adhoc.2025.104097

Mengdi Li, Zhuotong Feng, Bo Li

In emerging space–air–ground integrated networks, Low Earth Orbit (LEO) satellites offer edge computing capabilities. However, their limited onboard resources make efficient computation offloading a significant bottleneck. To address this challenge, we propose two novel algorithms that jointly optimize task offloading and resource allocation. The first is the Adaptive Large-Scale Task Priority Offloading and Resource Allocation (ALTPORA) algorithm, which prioritizes tasks to dynamically align offloading decisions with real-time satellite resource availability. The second is the Particle Swarm Optimization-based Joint Task Offloading and Resource Allocation (PSO-JTORA) algorithm, which employs a unified optimization framework to resolve resource contention. Simulation results demonstrate the superiority of our designs. ALTPORA consistently outperforms traditional methods in scalability and resource management. Notably, PSO-JTORA achieves up to 12.5% latency reduction compared with existing schemes and simultaneously lowers energy consumption under high-load conditions. These results highlight the effectiveness of our methods in balancing performance with resource constraints, providing a robust pathway toward more efficient and sustainable LEO satellite edge computing in future integrated networks.

在新兴的空-空-地综合网络中，低地球轨道卫星提供边缘计算能力。然而，它们有限的机载资源使得高效的计算卸载成为一个重要的瓶颈。为了解决这一挑战，我们提出了两种新的算法，共同优化任务卸载和资源分配。首先是自适应大规模任务优先级卸载和资源分配（ALTPORA）算法，该算法根据实时卫星资源可用性对任务进行优先级排序，从而动态调整卸载决策。二是基于粒子群优化的联合任务卸载和资源分配（PSO-JTORA）算法，该算法采用统一的优化框架解决资源争用问题。仿真结果证明了设计的优越性。ALTPORA在可伸缩性和资源管理方面始终优于传统方法。值得注意的是，与现有方案相比，PSO-JTORA实现了高达12.5%的延迟降低，同时降低了高负载条件下的能耗。这些结果突出了我们的方法在平衡性能和资源约束方面的有效性，为未来集成网络中更高效和可持续的LEO卫星边缘计算提供了强大的途径。

{"title":"Efficient joint optimization of task offloading and resource allocation for LEO satellite edge computing","authors":"Mengdi Li, Zhuotong Feng, Bo Li","doi":"10.1016/j.adhoc.2025.104097","DOIUrl":"10.1016/j.adhoc.2025.104097","url":null,"abstract":"<div><div>In emerging space–air–ground integrated networks, Low Earth Orbit (LEO) satellites offer edge computing capabilities. However, their limited onboard resources make efficient computation offloading a significant bottleneck. To address this challenge, we propose two novel algorithms that jointly optimize task offloading and resource allocation. The first is the Adaptive Large-Scale Task Priority Offloading and Resource Allocation (ALTPORA) algorithm, which prioritizes tasks to dynamically align offloading decisions with real-time satellite resource availability. The second is the Particle Swarm Optimization-based Joint Task Offloading and Resource Allocation (PSO-JTORA) algorithm, which employs a unified optimization framework to resolve resource contention. Simulation results demonstrate the superiority of our designs. ALTPORA consistently outperforms traditional methods in scalability and resource management. Notably, PSO-JTORA achieves up to 12.5% latency reduction compared with existing schemes and simultaneously lowers energy consumption under high-load conditions. These results highlight the effectiveness of our methods in balancing performance with resource constraints, providing a robust pathway toward more efficient and sustainable LEO satellite edge computing in future integrated networks.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"182 ","pages":"Article 104097"},"PeriodicalIF":4.8,"publicationDate":"2025-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145624180","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0