Pub Date : 2025-07-02DOI: 10.1016/j.vehcom.2025.100950
Wenjie Zhou, Tian Zhang, Zekun Lu, Linbo Zhai
As the Internet of Things (IoT) drives the development of Vehicular Edge Computing (VEC), there is a surge in computational demand from emerging in-vehicle applications. Most existing studies do not fully consider the frequent changes in network topology under high mobility of vehicles and the underutilization of idle resources by single-hop offloading. To this end, we propose a task offloading scheme for vehicular edge computing based on multi-hop offloading. The scheme allows task vehicles to offload tasks to service vehicles with excess idle resources outside the communication range, and adapts to dynamic changes in network topology by introducing the concept of neighboring vehicle connection time. This study aims to minimize the delayed energy consumption utility value of the task under the conditions of satisfying the maximum task delay limit, vehicle computational and storage resource constraints. In response to this NP-hard problem, a two-stage reinforcement learning strategy MOCDD (combining Deep Q Network (DQN) and Deep Deterministic Policy Gradient (DDPG)) is proposed to divide the mixed action space into pure discrete and pure continuous action space to determine task migration, executive decision and vehicle transmission power. Simulation results verify the effectiveness of the proposed scheme.
{"title":"Deep reinforcement learning based migration and execution decisions for multi-hop task offloading in mobile vehicle edge computing","authors":"Wenjie Zhou, Tian Zhang, Zekun Lu, Linbo Zhai","doi":"10.1016/j.vehcom.2025.100950","DOIUrl":"10.1016/j.vehcom.2025.100950","url":null,"abstract":"<div><div>As the Internet of Things (IoT) drives the development of Vehicular Edge Computing (VEC), there is a surge in computational demand from emerging in-vehicle applications. Most existing studies do not fully consider the frequent changes in network topology under high mobility of vehicles and the underutilization of idle resources by single-hop offloading. To this end, we propose a task offloading scheme for vehicular edge computing based on multi-hop offloading. The scheme allows task vehicles to offload tasks to service vehicles with excess idle resources outside the communication range, and adapts to dynamic changes in network topology by introducing the concept of neighboring vehicle connection time. This study aims to minimize the delayed energy consumption utility value of the task under the conditions of satisfying the maximum task delay limit, vehicle computational and storage resource constraints. In response to this NP-hard problem, a two-stage reinforcement learning strategy MOCDD (combining Deep Q Network (DQN) and Deep Deterministic Policy Gradient (DDPG)) is proposed to divide the mixed action space into pure discrete and pure continuous action space to determine task migration, executive decision and vehicle transmission power. Simulation results verify the effectiveness of the proposed scheme.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"55 ","pages":"Article 100950"},"PeriodicalIF":5.8,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144563853","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-25DOI: 10.1016/j.vehcom.2025.100948
Mingyu Zhang , Zhibo Sun , Fengjie Li , Hong Zhang
With the rapid growth of Internet of Things (IoT) devices, Mobile Edge Computing (MEC) faces challenges in meeting increasing computational demands, especially in resource-constrained environments. To address this issue, we propose the LEO Satellite-MS-RSU Edge Computing (LMREC) framework, which integrates Mobile Servers (MSs), Low Earth Orbit (LEO) satellite networks, and Roadside Units (RSUs) into an innovative edge computing architecture. We first introduce “attraction” and “repulsion” metrics to model the willingness of vehicular satellite servers to serve specific users. Subsequently, we design a Magnetic Equilibrium Algorithm (MEA), which dynamically adjusts the MS deployment and service allocation by balancing user-driven attraction and server repulsion. To address the latency sensitivity of task scheduling and user satisfaction in LMREC, we formulate a mixed-integer nonlinear programming (MINLP) optimization problem for task offloading and resource allocation. Since this optimization problem is intractable to solve in polynomial time, we propose a Magnetic Domain Migration Algorithm (MDMA) to obtain a near-optimal solution. In MDMA, tasks are modeled as magnetic domains migrating in a magnetic field, and the optimization problem is decomposed into subproblems, which are solved using Exact Potential Game Theory, convex optimization, and a hybrid genetic algorithm. Finally, simulation results validate the effectiveness of the LMREC framework, demonstrating its superiority over existing methods and its potential to enhance collaboration among end devices, RSUs, and LEO satellite networks.
随着物联网（IoT）设备的快速增长，移动边缘计算（MEC）在满足日益增长的计算需求方面面临挑战，特别是在资源受限的环境下。为了解决这一问题，我们提出了低轨道卫星- ms - rsu边缘计算（LMREC）框架，该框架将移动服务器（ms）、低地球轨道（LEO）卫星网络和路边单元（rsu）集成到一个创新的边缘计算架构中。我们首先引入“吸引力”和“排斥力”指标来模拟车载卫星服务器为特定用户服务的意愿。随后，我们设计了一个磁平衡算法（MEA），该算法通过平衡用户驱动的吸引力和服务器排斥来动态调整MS的部署和服务分配。为了解决任务调度的延迟敏感性和用户满意度问题，我们提出了一个混合整数非线性规划（MINLP）的任务卸载和资源分配优化问题。由于该优化问题难以在多项式时间内解决，我们提出了一种磁域迁移算法（MDMA）来获得近似最优解。在MDMA中，将任务建模为磁场中的磁域迁移，并将优化问题分解为子问题，利用精确势博弈论、凸优化和混合遗传算法求解子问题。最后，仿真结果验证了LMREC框架的有效性，表明其优于现有方法，并具有增强终端设备、rsu和LEO卫星网络之间协作的潜力。
{"title":"MIEC: A magnetism-inspired framework for MS deployment and joint task offloading and resource allocation optimization in LMREC","authors":"Mingyu Zhang , Zhibo Sun , Fengjie Li , Hong Zhang","doi":"10.1016/j.vehcom.2025.100948","DOIUrl":"10.1016/j.vehcom.2025.100948","url":null,"abstract":"<div><div>With the rapid growth of Internet of Things (IoT) devices, Mobile Edge Computing (MEC) faces challenges in meeting increasing computational demands, especially in resource-constrained environments. To address this issue, we propose the LEO Satellite-MS-RSU Edge Computing (LMREC) framework, which integrates Mobile Servers (MSs), Low Earth Orbit (LEO) satellite networks, and Roadside Units (RSUs) into an innovative edge computing architecture. We first introduce “attraction” and “repulsion” metrics to model the willingness of vehicular satellite servers to serve specific users. Subsequently, we design a Magnetic Equilibrium Algorithm (MEA), which dynamically adjusts the MS deployment and service allocation by balancing user-driven attraction and server repulsion. To address the latency sensitivity of task scheduling and user satisfaction in LMREC, we formulate a mixed-integer nonlinear programming (MINLP) optimization problem for task offloading and resource allocation. Since this optimization problem is intractable to solve in polynomial time, we propose a Magnetic Domain Migration Algorithm (MDMA) to obtain a near-optimal solution. In MDMA, tasks are modeled as magnetic domains migrating in a magnetic field, and the optimization problem is decomposed into subproblems, which are solved using Exact Potential Game Theory, convex optimization, and a hybrid genetic algorithm. Finally, simulation results validate the effectiveness of the LMREC framework, demonstrating its superiority over existing methods and its potential to enhance collaboration among end devices, RSUs, and LEO satellite networks.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"55 ","pages":"Article 100948"},"PeriodicalIF":5.8,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144491512","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-25DOI: 10.1016/j.vehcom.2025.100949
Oluwatosin Ahmed Amodu , Huda Althumali , Zurina Mohd Hanapi , Chedia Jarray , Raja Azlina Raja Mahmood , Mohammed Sani Adam , Umar Ali Bukar , Nor Fadzilah Abdullah , Nguyen Cong Luong
Unmanned Aerial Vehicles (UAVs) play a critical role in data collection for a wide range of Internet of Things (IoT) applications across remote, urban, and marine environments. In large-scale deployments, UAVs often face complex decision-making challenges, for which Deep Reinforcement Learning (DRL) has emerged as a promising solution. This paper presents a comprehensive review of research on UAV-assisted IoT utilizing DRL, covering key research questions relating to DRL algorithm variants, deployment objectives, architectural features, integrated technologies, UAV roles, optimization constraints, energy management strategies, and performance metrics. Findings indicate that value-based and actor-critic algorithms are the most commonly employed, targeting objectives such as path planning, transmit power control, scheduling, velocity and altitude control, and charging optimization. Other architectural considerations include clustering, security, obstacle avoidance, buffered sensors, and multi-UAV coordination. Beyond data collection, UAVs are also used for tasks such as device selection, data aggregation, and sensor charging, with energy management primarily achieved through charging and energy harvesting techniques. Performance is typically assessed using metrics like energy efficiency, throughput, latency, packet loss, and Age of Information (AoI). The paper concludes by outlining several promising research directions and open challenges critical to the successful deployment of UAVs as aerial communication platforms, especially in IoT data collection. By organizing existing work across key themes and outlining promising future directions, this review offers a valuable reference for researchers and technology professionals alike.
无人机（uav）在远程、城市和海洋环境中广泛的物联网（IoT）应用的数据收集中发挥着关键作用。在大规模部署中，无人机经常面临复杂的决策挑战，深度强化学习（DRL）已成为一种有前途的解决方案。本文全面回顾了利用DRL的无人机辅助物联网应用，涵盖了关键研究问题、DRL算法变体、部署目标、架构特征、集成技术、无人机角色、优化约束、能源管理策略和性能指标。研究结果表明，基于价值的算法和行为者批评算法是最常用的，针对的目标包括路径规划、传输功率控制、调度、速度和高度控制以及充电优化。架构考虑包括集群、安全性、避障、缓冲传感器和多无人机协调。除了数据收集，无人机还用于设备选择、数据聚合和传感器充电等任务，主要通过充电和能量收集技术实现能量管理。通常使用能效、吞吐量、延迟、数据包丢失和信息年龄（Age of Information, AoI）等指标来评估性能。最后，本文概述了几个有前途的研究方向和开放的挑战，这些挑战对无人机在物联网数据收集中的成功部署至关重要。
{"title":"A comprehensive survey of deep reinforcement learning in UAV-assisted IoT data collection","authors":"Oluwatosin Ahmed Amodu , Huda Althumali , Zurina Mohd Hanapi , Chedia Jarray , Raja Azlina Raja Mahmood , Mohammed Sani Adam , Umar Ali Bukar , Nor Fadzilah Abdullah , Nguyen Cong Luong","doi":"10.1016/j.vehcom.2025.100949","DOIUrl":"10.1016/j.vehcom.2025.100949","url":null,"abstract":"<div><div>Unmanned Aerial Vehicles (UAVs) play a critical role in data collection for a wide range of Internet of Things (IoT) applications across remote, urban, and marine environments. In large-scale deployments, UAVs often face complex decision-making challenges, for which Deep Reinforcement Learning (DRL) has emerged as a promising solution. This paper presents a comprehensive review of research on UAV-assisted IoT utilizing DRL, covering key research questions relating to DRL algorithm variants, deployment objectives, architectural features, integrated technologies, UAV roles, optimization constraints, energy management strategies, and performance metrics. Findings indicate that value-based and actor-critic algorithms are the most commonly employed, targeting objectives such as path planning, transmit power control, scheduling, velocity and altitude control, and charging optimization. Other architectural considerations include clustering, security, obstacle avoidance, buffered sensors, and multi-UAV coordination. Beyond data collection, UAVs are also used for tasks such as device selection, data aggregation, and sensor charging, with energy management primarily achieved through charging and energy harvesting techniques. Performance is typically assessed using metrics like energy efficiency, throughput, latency, packet loss, and Age of Information (AoI). The paper concludes by outlining several promising research directions and open challenges critical to the successful deployment of UAVs as aerial communication platforms, especially in IoT data collection. By organizing existing work across key themes and outlining promising future directions, this review offers a valuable reference for researchers and technology professionals alike.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"55 ","pages":"Article 100949"},"PeriodicalIF":5.8,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144515946","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-23DOI: 10.1016/j.vehcom.2025.100947
Chen Sun, Liang Hou, Suqi Yu, Jian Shu
Reliable and efficient data transmission between Unmanned Aerial Vehicle (UAV) nodes is critical for the control of UAV swarms and relies heavily on effective routing protocols in Flying Ad-hoc Networks (FANETs). However, Q-learning-based FANET routing protocols, which are gaining widespread attention, face two significant challenges: 1) insufficient stability of Q-learning leads to unreliable route selection in certain scenarios and higher packet loss rates; 2) in void regions with frequent topology changes and vast path exploration spaces, the slow convergence of Q-learning fails to adapt quickly to dynamic environmental changes, thereby reducing the packet delivery rate (PDR). This paper proposes a hybrid Q-learning/AODV (HQA) multi-path routing algorithm that integrates Q-learning and the AODV protocols to address these challenges. HQA includes a Bayesian stability evaluator for adaptive Q-learning/AODV switching and a dual-update reward mechanism that integrates reliable AODV paths into Q-learning training, enabling rapid void recovery and latency-optimized routing. Experimental results demonstrate HQA's superiority over baseline protocols: Compared to AODV, HQA reduces average end-to-end delay by 13.6–23.9% and improves PDR by 5.4–9.1% in non-void and void states, respectively. It outperforms QMR by 2.2–6.3% in PDR while achieving 25.6% and 53.2% higher average PDR than QMR and AODV across network densities. The hybrid design accelerates convergence by 40% versus standalone Q-learning through AODV-assisted rewards, maintaining scalability under dynamic topology changes. These findings indicate that the HQA algorithm can more rapidly adapt to the rapid changes in FANETs and better handle void regions, offering a promising solution for enhancing the performance and reliability of FANETs.
{"title":"HQA: Hybrid Q-learning and AODV multi-path routing algorithm for Flying Ad-hoc Networks","authors":"Chen Sun, Liang Hou, Suqi Yu, Jian Shu","doi":"10.1016/j.vehcom.2025.100947","DOIUrl":"10.1016/j.vehcom.2025.100947","url":null,"abstract":"<div><div>Reliable and efficient data transmission between Unmanned Aerial Vehicle (UAV) nodes is critical for the control of UAV swarms and relies heavily on effective routing protocols in Flying Ad-hoc Networks (FANETs). However, Q-learning-based FANET routing protocols, which are gaining widespread attention, face two significant challenges: 1) insufficient stability of Q-learning leads to unreliable route selection in certain scenarios and higher packet loss rates; 2) in void regions with frequent topology changes and vast path exploration spaces, the slow convergence of Q-learning fails to adapt quickly to dynamic environmental changes, thereby reducing the packet delivery rate (PDR). This paper proposes a hybrid Q-learning/AODV (HQA) multi-path routing algorithm that integrates Q-learning and the AODV protocols to address these challenges. HQA includes a Bayesian stability evaluator for adaptive Q-learning/AODV switching and a dual-update reward mechanism that integrates reliable AODV paths into Q-learning training, enabling rapid void recovery and latency-optimized routing. Experimental results demonstrate HQA's superiority over baseline protocols: Compared to AODV, HQA reduces average end-to-end delay by 13.6–23.9% and improves PDR by 5.4–9.1% in non-void and void states, respectively. It outperforms QMR by 2.2–6.3% in PDR while achieving 25.6% and 53.2% higher average PDR than QMR and AODV across network densities. The hybrid design accelerates convergence by 40% versus standalone Q-learning through AODV-assisted rewards, maintaining scalability under dynamic topology changes. These findings indicate that the HQA algorithm can more rapidly adapt to the rapid changes in FANETs and better handle void regions, offering a promising solution for enhancing the performance and reliability of FANETs.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"55 ","pages":"Article 100947"},"PeriodicalIF":5.8,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144491511","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-20DOI: 10.1016/j.vehcom.2025.100946
Mohammed A. Abdelmaguid, Hossam S. Hassanein, Mohammad Zulkernine
Addressing known threats constitutes the foundational layer of cybersecurity defenses. However, the real challenge emerges in anticipating and mitigating unforeseen attacks. Current security methodologies work well against familiar threats but often struggle with new or unforeseen attacks. This paper examines the Trust Origin within Trust Management Systems (TMS) by linking it to the network attack rate, thereby refining trust assessments and predicting new attacks. Combining Machine Learning (ML) algorithms with honeypots, we offer a comprehensive defense for Vehicular Ad-hoc Networks (VANETs), adept at detecting anticipated and unexpected attacks through attack rate analysis. Our methodology evaluates the network's security status by examining its ability to identify known attacks, referred to as prepared-for attacks. Subsequently, this information serves as a foundation to predict future attacks that still need to be identified, termed unprepared-for attacks. Through extensive testing, we demonstrate the viability of a dual strategy that encompasses the detection of prepared-for attacks and the prediction of unprepared-for ones. Experimental results reveal a significant improvement in predicting unprepared-for attacks, evidenced by enhanced accuracy, precision, and recall. Additionally, we conduct experiments to determine the optimal deployment of honeypots for maximum efficiency.
{"title":"Securing the unforeseen: Enhancing VANET security with dynamic honeypots and attack rate analysis","authors":"Mohammed A. Abdelmaguid, Hossam S. Hassanein, Mohammad Zulkernine","doi":"10.1016/j.vehcom.2025.100946","DOIUrl":"10.1016/j.vehcom.2025.100946","url":null,"abstract":"<div><div>Addressing known threats constitutes the foundational layer of cybersecurity defenses. However, the real challenge emerges in anticipating and mitigating unforeseen attacks. Current security methodologies work well against familiar threats but often struggle with new or unforeseen attacks. This paper examines the Trust Origin within Trust Management Systems (TMS) by linking it to the network attack rate, thereby refining trust assessments and predicting new attacks. Combining Machine Learning (ML) algorithms with honeypots, we offer a comprehensive defense for Vehicular Ad-hoc Networks (VANETs), adept at detecting anticipated and unexpected attacks through attack rate analysis. Our methodology evaluates the network's security status by examining its ability to identify known attacks, referred to as prepared-for attacks. Subsequently, this information serves as a foundation to predict future attacks that still need to be identified, termed unprepared-for attacks. Through extensive testing, we demonstrate the viability of a dual strategy that encompasses the detection of prepared-for attacks and the prediction of unprepared-for ones. Experimental results reveal a significant improvement in predicting unprepared-for attacks, evidenced by enhanced accuracy, precision, and recall. Additionally, we conduct experiments to determine the optimal deployment of honeypots for maximum efficiency.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"55 ","pages":"Article 100946"},"PeriodicalIF":5.8,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144337779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-06DOI: 10.1016/j.vehcom.2025.100945
Peng Chen , Ting Zhou , Zhimin Chen , Fan Meng , Jun Liu
To enable the next generation of connected autonomous vehicles, the millimeter wave (mmWave)-based integrated sensing and communication (ISAC) system will be a critical technology in future vehicle-to-everything (V2X) networks. However, the rapid mobility of vehicles and the narrow beamwidth of mmWave signals present significant challenges for beam alignment, and point-target modeling methods often lead to substantial overhead, high latency, and complications. To address these issues, in this paper, a hybrid analog-digital (HAD) multi-input multi-output (MIMO) ISAC framework is adopted for the mmWave-based V2X network to reduce hardware costs and power consumption. Then, considering the narrow beamwidth of the mmWave system, the vehicle is modeled as an extended surface target with multiple scattering points, and a new association technique for these points is developed to improve prediction accuracy. Hence, a deep learning (DL)-based beamforming prediction network, namely beamforming prediction network (BFP-Net), is designed according to the ISAC signal beam prediction protocol and enables roadside units (RSUs) to transmit ISAC signals effectively for both downlink communication and sensing operations. The BFP-Net leverages a convolutional neural network long-short-term memory (CNN-LSTM) architecture to capture spatial and temporal correlations, providing enhanced modeling capabilities for beam prediction. Moreover, for highly dynamic vehicles, the BFP-Net predicts optimal beams for future time slots by extracting features from the received echo signals and eliminates the repetitive beam training inherent in the traditional communication protocol. Simulation results demonstrate that the proposed method significantly outperforms extended Kalman filter (EKF)-based methods in the mmWave V2X scenario, achieving higher beam gains and better performance for high-speed vehicles, and substantially reduces the overhead associated with beam training compared to the conventional neural network relying on pilot signals.
{"title":"BFP-Net: A DL-based ISAC beamforming prediction method for extended vehicle","authors":"Peng Chen , Ting Zhou , Zhimin Chen , Fan Meng , Jun Liu","doi":"10.1016/j.vehcom.2025.100945","DOIUrl":"10.1016/j.vehcom.2025.100945","url":null,"abstract":"<div><div>To enable the next generation of connected autonomous vehicles, the millimeter wave (mmWave)-based integrated sensing and communication (ISAC) system will be a critical technology in future vehicle-to-everything (V2X) networks. However, the rapid mobility of vehicles and the narrow beamwidth of mmWave signals present significant challenges for beam alignment, and point-target modeling methods often lead to substantial overhead, high latency, and complications. To address these issues, in this paper, a hybrid analog-digital (HAD) multi-input multi-output (MIMO) ISAC framework is adopted for the mmWave-based V2X network to reduce hardware costs and power consumption. Then, considering the narrow beamwidth of the mmWave system, the vehicle is modeled as an extended surface target with multiple scattering points, and a new association technique for these points is developed to improve prediction accuracy. Hence, a deep learning (DL)-based beamforming prediction network, namely beamforming prediction network (BFP-Net), is designed according to the ISAC signal beam prediction protocol and enables roadside units (RSUs) to transmit ISAC signals effectively for both downlink communication and sensing operations. The BFP-Net leverages a convolutional neural network long-short-term memory (CNN-LSTM) architecture to capture spatial and temporal correlations, providing enhanced modeling capabilities for beam prediction. Moreover, for highly dynamic vehicles, the BFP-Net predicts optimal beams for future time slots by extracting features from the received echo signals and eliminates the repetitive beam training inherent in the traditional communication protocol. Simulation results demonstrate that the proposed method significantly outperforms extended Kalman filter (EKF)-based methods in the mmWave V2X scenario, achieving higher beam gains and better performance for high-speed vehicles, and substantially reduces the overhead associated with beam training compared to the conventional neural network relying on pilot signals.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"55 ","pages":"Article 100945"},"PeriodicalIF":5.8,"publicationDate":"2025-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243172","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}
Flexible movement and rapid deployment capabilities of unmanned aerial vehicles (UAVs) or drones have enabled them to be ideal for fresh and real-time data collection in the Internet of Drones (IoD) network. With the rising demand for IoD applications, optimizing the Age of Information (AoI), and energy efficiency of drones has become a challenging problem. The existing literature works are either limited by considering single-drone data collection from 2D space or by not prioritizing data from diverse IoT devices. In this paper, we have developed an optimization framework for multi-drone-assisted data collection in 3D space, which brings a trade-off between minimizing drone energy consumption and AoI, exploiting the Mixed Integer Linear Programming (MILP) problem. However, due to the NP-hardness of the developed optimization framework for large networks, we have devised a Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm, supported and enhanced by a Multi-Head Attention (MHA) mechanism for multi-drone-assisted data collection to minimize drone energy consumption and AoI jointly, namely MECAO. The MHA in the MECAO system helps prioritize IoT data sources and ensures the timely collection of important data. This system enables the agents to coordinate effectively among themselves and provides innovative solutions to complex network issues. Our findings demonstrate substantial advancements in real-time data collection and drone performance, offering a practical and efficient solution for modern IoD applications. The developed MECAO system is implemented in the OpenAI Gym simulator platform, and the simulation trace file content depicts the improvement in AoI by up to 56% while the energy consumption is reduced by as high as 38.5%, respectively, compared to the state-of-the-art works.
{"title":"A Multi-Head Attention mechanism assisted MADDPG algorithm for real-time data collection in Internet of Drones","authors":"A.K.M. Atiqur Rahman , Muntasir Chowdhury Mridul , Palash Roy , Md. Abdur Razzaque , Md. Rajin Saleh , Mohammad Mehedi Hassan , Md Zia Uddin","doi":"10.1016/j.vehcom.2025.100944","DOIUrl":"10.1016/j.vehcom.2025.100944","url":null,"abstract":"<div><div>Flexible movement and rapid deployment capabilities of unmanned aerial vehicles (UAVs) or drones have enabled them to be ideal for fresh and real-time data collection in the Internet of Drones (IoD) network. With the rising demand for IoD applications, optimizing the Age of Information (AoI), and energy efficiency of drones has become a challenging problem. The existing literature works are either limited by considering single-drone data collection from 2D space or by not prioritizing data from diverse IoT devices. In this paper, we have developed an optimization framework for multi-drone-assisted data collection in 3D space, which brings a trade-off between minimizing drone energy consumption and AoI, exploiting the Mixed Integer Linear Programming (MILP) problem. However, due to the NP-hardness of the developed optimization framework for large networks, we have devised a Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm, supported and enhanced by a Multi-Head Attention (MHA) mechanism for multi-drone-assisted data collection to minimize drone energy consumption and AoI jointly, namely MECAO. The MHA in the MECAO system helps prioritize IoT data sources and ensures the timely collection of important data. This system enables the agents to coordinate effectively among themselves and provides innovative solutions to complex network issues. Our findings demonstrate substantial advancements in real-time data collection and drone performance, offering a practical and efficient solution for modern IoD applications. The developed MECAO system is implemented in the OpenAI Gym simulator platform, and the simulation trace file content depicts the improvement in AoI by up to 56% while the energy consumption is reduced by as high as 38.5%, respectively, compared to the state-of-the-art works.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"54 ","pages":"Article 100944"},"PeriodicalIF":5.8,"publicationDate":"2025-05-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144205001","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-05-26DOI: 10.1016/j.vehcom.2025.100941
Can Wang, Junhong Zhang, Helin Yang
The cognitive unmanned aerial vehicle (UAV) communication system has emerged as a pivotal technology in addressing the scarcity of spectral resources for UAV communications, but the jamming and eavesdropping attacks are severe due to the high-quality air-to-ground communication links. Consequently, this paper introduces a UAV-enabled cooperative jammer to disrupt the eavesdropping activities of active eavesdroppers by emitting artificial noise. Our objective is to jointly optimize the three-dimensional UAV trajectory and transmit power to maximize the secrecy communication rate under quality of service (QoS) requirement. To tackle the non-convex problem, the block coordinate descent (BCD) and successive convex approximation (SCA) methods are utilized to transform it into an approximate convex problem, and then we design an alternative optimization iterative algorithm to achieve suboptimal but efficient solution. Moreover, we extend the developed algorithm into an imperfect channel state information (CSI) scenario to maximize the worst-case secrecy rate by jointly optimizing the robust UAV's trajectory and transmit power, where the location uncertainties of ground primary, secondary, and eavesdropping devices are considered. Simulation results demonstrate that the proposed joint optimization algorithm significantly enhances system secrecy performance under different real-world settings compared to existing state-of-the-art algorithms.
{"title":"Cognitive UAV-assisted secure and reliable communications based on robust joint trajectory and power control optimization","authors":"Can Wang, Junhong Zhang, Helin Yang","doi":"10.1016/j.vehcom.2025.100941","DOIUrl":"10.1016/j.vehcom.2025.100941","url":null,"abstract":"<div><div>The cognitive unmanned aerial vehicle (UAV) communication system has emerged as a pivotal technology in addressing the scarcity of spectral resources for UAV communications, but the jamming and eavesdropping attacks are severe due to the high-quality air-to-ground communication links. Consequently, this paper introduces a UAV-enabled cooperative jammer to disrupt the eavesdropping activities of active eavesdroppers by emitting artificial noise. Our objective is to jointly optimize the three-dimensional UAV trajectory and transmit power to maximize the secrecy communication rate under quality of service (QoS) requirement. To tackle the non-convex problem, the block coordinate descent (BCD) and successive convex approximation (SCA) methods are utilized to transform it into an approximate convex problem, and then we design an alternative optimization iterative algorithm to achieve suboptimal but efficient solution. Moreover, we extend the developed algorithm into an imperfect channel state information (CSI) scenario to maximize the worst-case secrecy rate by jointly optimizing the robust UAV's trajectory and transmit power, where the location uncertainties of ground primary, secondary, and eavesdropping devices are considered. Simulation results demonstrate that the proposed joint optimization algorithm significantly enhances system secrecy performance under different real-world settings compared to existing state-of-the-art algorithms.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"54 ","pages":"Article 100941"},"PeriodicalIF":5.8,"publicationDate":"2025-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144170397","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-05-23DOI: 10.1016/j.vehcom.2025.100940
Xiangrui Guan, Jianbin Xue, Han Zhang, Jialing Xu
The combination of the intelligent reflecting surface (IRS) with reconfigurable wireless propagation environment and the millimeter-wave (mmWave) with abundant bandwidth resources can play a great advantage over the rate and delay in vehicular communications. Considering the problem of non-line-of-sight (NLOS) communication between the requesting nodes (RNs) and the service nodes (SNs) in the mmWave vehicular system in this paper, we propose an IRS-assisted multi-hop vehicle-to-vehicle (V2V) cooperative communication method to realize low-delay vehicular communication. Aiming to minimize the communication delay of RNs, an optimization problem is formulated by optimizing the link selection and reflection coefficient matrix of IRS. To tackle the optimization problem, an alternate optimization algorithm is proposed to decompose the original optimization problem into two subproblems for iterative optimization. First, we establish a link selection mechanism based on link quality and vehicle distance and propose a link selection algorithm based on the evaluation function to select communication links for each RN. Then, in particular, we derive the closed-form expression based on successive convex approximation (SCA) techniques for updating the reflection coefficient matrix of IRS. The simulation results show that the IRS-assisted mmWave vehicular cooperative communication scheme proposed in this paper can effectively reduce the communication delay and improve the performance of the mmWave vehicular network.
{"title":"Millimeter-wave vehicular collaborative communication assisted by intelligent reflecting surface","authors":"Xiangrui Guan, Jianbin Xue, Han Zhang, Jialing Xu","doi":"10.1016/j.vehcom.2025.100940","DOIUrl":"10.1016/j.vehcom.2025.100940","url":null,"abstract":"<div><div>The combination of the intelligent reflecting surface (IRS) with reconfigurable wireless propagation environment and the millimeter-wave (mmWave) with abundant bandwidth resources can play a great advantage over the rate and delay in vehicular communications. Considering the problem of non-line-of-sight (NLOS) communication between the requesting nodes (RNs) and the service nodes (SNs) in the mmWave vehicular system in this paper, we propose an IRS-assisted multi-hop vehicle-to-vehicle (V2V) cooperative communication method to realize low-delay vehicular communication. Aiming to minimize the communication delay of RNs, an optimization problem is formulated by optimizing the link selection and reflection coefficient matrix of IRS. To tackle the optimization problem, an alternate optimization algorithm is proposed to decompose the original optimization problem into two subproblems for iterative optimization. First, we establish a link selection mechanism based on link quality and vehicle distance and propose a link selection algorithm based on the evaluation function to select communication links for each RN. Then, in particular, we derive the closed-form expression based on successive convex approximation (SCA) techniques for updating the reflection coefficient matrix of IRS. The simulation results show that the IRS-assisted mmWave vehicular cooperative communication scheme proposed in this paper can effectively reduce the communication delay and improve the performance of the mmWave vehicular network.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"54 ","pages":"Article 100940"},"PeriodicalIF":5.8,"publicationDate":"2025-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144195605","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-05-22DOI: 10.1016/j.vehcom.2025.100942
Suhui Liu , Liquan Chen , Liqun Chen , Yu Wang , Yaqing Zhu
Vehicle-to-infrastructure (V2I) communication is the basis for vehicles to obtain information about the road ahead. The confidentiality and reliability of V2I communication guarantee traffic safety and smooth flow. Authenticated key agreement (AKA) is the most commonly used technique to establish secure communication channels. Signature-based AKA inevitably exposes the identity information of vehicles, while Encryption-based AKA can bring deniability and high privacy, which means no adversary can know who sent the AKA message. Certificateless encryption (CLE) can simultaneously solve burdensome certificate management and key escrow. However, existing certificateless cryptography requires two loosely combined public keys to represent a device and does not consider the physical security of storing secret keys locally. This paper first designed an improved CLE scheme with one-device-one-public-key, and performance comparisons show that the proposed CLE has optimal storage and computation performance. Considering that rare work was put on encryption-based AKA, this paper proposed a deniable and privacy-preserving certificateless AKA for V2I communication by incorporating Physically Unclonable Function (PUF)-secured key management to prevent physical leakage of keys, named CLE-AKA-PUF. Feature comparison illustrates that CLE-AKA-PUF supports key escrow-free, dual authentication, physical security, deniability, and high privacy. Security proofs and performance analysis demonstrate the practicability and efficiency of CLE-AKA-PUF.
{"title":"CLE-based authenticated key agreement with PUF-secured key for vehicle-to-infrastructure","authors":"Suhui Liu , Liquan Chen , Liqun Chen , Yu Wang , Yaqing Zhu","doi":"10.1016/j.vehcom.2025.100942","DOIUrl":"10.1016/j.vehcom.2025.100942","url":null,"abstract":"<div><div>Vehicle-to-infrastructure (V2I) communication is the basis for vehicles to obtain information about the road ahead. The confidentiality and reliability of V2I communication guarantee traffic safety and smooth flow. Authenticated key agreement (AKA) is the most commonly used technique to establish secure communication channels. Signature-based AKA inevitably exposes the identity information of vehicles, while Encryption-based AKA can bring deniability and high privacy, which means no adversary can know who sent the AKA message. Certificateless encryption (CLE) can simultaneously solve burdensome certificate management and key escrow. However, existing certificateless cryptography requires two loosely combined public keys to represent a device and does not consider the physical security of storing secret keys locally. This paper first designed an improved CLE scheme with one-device-one-public-key, and performance comparisons show that the proposed CLE has optimal storage and computation performance. Considering that rare work was put on encryption-based AKA, this paper proposed a deniable and privacy-preserving certificateless AKA for V2I communication by incorporating Physically Unclonable Function (PUF)-secured key management to prevent physical leakage of keys, named CLE-AKA-PUF. Feature comparison illustrates that CLE-AKA-PUF supports key escrow-free, dual authentication, physical security, deniability, and high privacy. Security proofs and performance analysis demonstrate the practicability and efficiency of CLE-AKA-PUF.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"54 ","pages":"Article 100942"},"PeriodicalIF":5.8,"publicationDate":"2025-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144170398","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号