In vehicular ad-hoc networks (VANETs), ensuring robust security for vehicle identities and messages while maintaining essential service functionalities presents a significant challenge. This paper proposes a group signature-based anonymous authentication scheme for VANETs (GSAAS). GSAAS supports anonymous vehicle authentication within a certificateless framework, effectively mitigating the complexities associated with certificate management and distribution. To alleviate the high computation overhead on the Trust Authority (TA) and minimize the communication delay associated with pseudonym requests, the base station (BS) is employed as the group manager, enabling efficient group maintenance and pseudonym management, facilitating seamless vehicle authentication while ensuring secure data transmission. Security analysis demonstrates that GSAAS is robust against various attacks. Furthermore, performance analysis highlights the superior efficiency of GSAAS compared to existing schemes, with significant improvements in both computation and communication overheads in VANETs.
{"title":"GSAAS: A group signature-based anonymous authentication scheme for VANETs","authors":"Xinyang Deng , Xiaohong Wu , Qinggele Qi , Cong Zhao","doi":"10.1016/j.vehcom.2025.100988","DOIUrl":"10.1016/j.vehcom.2025.100988","url":null,"abstract":"<div><div>In vehicular ad-hoc networks (VANETs), ensuring robust security for vehicle identities and messages while maintaining essential service functionalities presents a significant challenge. This paper proposes a group signature-based anonymous authentication scheme for VANETs (GSAAS). GSAAS supports anonymous vehicle authentication within a certificateless framework, effectively mitigating the complexities associated with certificate management and distribution. To alleviate the high computation overhead on the Trust Authority (TA) and minimize the communication delay associated with pseudonym requests, the base station (BS) is employed as the group manager, enabling efficient group maintenance and pseudonym management, facilitating seamless vehicle authentication while ensuring secure data transmission. Security analysis demonstrates that GSAAS is robust against various attacks. Furthermore, performance analysis highlights the superior efficiency of GSAAS compared to existing schemes, with significant improvements in both computation and communication overheads in VANETs.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"57 ","pages":"Article 100988"},"PeriodicalIF":6.5,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145498831","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 : 2025-11-09DOI: 10.1016/j.vehcom.2025.100989
Huayu Liu , Hua Wu , Yang Liu , Hailong Dong , Hao Li
Multiple unmanned aerial vehicles (UAVs) play a critical role in disaster response and rescue missions. This paper proposes a multi-layer based collaborative optimization (MCO) method, which consists of three stages: path preplanning, task allocation, and task scheduling. These three stages correspond to three levels that are upper level, middle level and lower level. A dynamic constrained particle swarm optimization (DPSO) is proposed for path preplanning in the upper layer by designing a kind of dynamic subpopulation division strategy. After that a clustered consensus-based bundle algorithm (CCBA) is designed to allocate different tasks to available UAVs based on preplanned paths to solve the problems of discontinuous task allocation and redundant paths. Then a multi-neighborhood variable simulated annealing (MNV-SA) algorithm is proposed to further optimize the task execution sequence of each UAV. To validate the effectiveness of MCO method, a set of experiments is conducted in a simulated disaster scenario based on a real urban environment. The results demonstrate that the proposed MCO method significantly improves the task execution benefits and reduces UAV flight distances across all scenarios. Particularly, in complex scenarios, MCO method outperforms CBBA, ACO, and PI algorithms in terms of task execution benefits by 14.01 %, 6.01 %, and 24.06 %, respectively.
{"title":"A multi-layer based collaborative optimization (MCO) for multiple UAVs’ task allocation and scheduling","authors":"Huayu Liu , Hua Wu , Yang Liu , Hailong Dong , Hao Li","doi":"10.1016/j.vehcom.2025.100989","DOIUrl":"10.1016/j.vehcom.2025.100989","url":null,"abstract":"<div><div>Multiple unmanned aerial vehicles (UAVs) play a critical role in disaster response and rescue missions. This paper proposes a multi-layer based collaborative optimization (MCO) method, which consists of three stages: path preplanning, task allocation, and task scheduling. These three stages correspond to three levels that are upper level, middle level and lower level. A dynamic constrained particle swarm optimization (DPSO) is proposed for path preplanning in the upper layer by designing a kind of dynamic subpopulation division strategy. After that a clustered consensus-based bundle algorithm (CCBA) is designed to allocate different tasks to available UAVs based on preplanned paths to solve the problems of discontinuous task allocation and redundant paths. Then a multi-neighborhood variable simulated annealing (MNV-SA) algorithm is proposed to further optimize the task execution sequence of each UAV. To validate the effectiveness of MCO method, a set of experiments is conducted in a simulated disaster scenario based on a real urban environment. The results demonstrate that the proposed MCO method significantly improves the task execution benefits and reduces UAV flight distances across all scenarios. Particularly, in complex scenarios, MCO method outperforms CBBA, ACO, and PI algorithms in terms of task execution benefits by 14.01 %, 6.01 %, and 24.06 %, respectively.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"57 ","pages":"Article 100989"},"PeriodicalIF":6.5,"publicationDate":"2025-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145473283","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 : 2025-11-07DOI: 10.1016/j.vehcom.2025.100984
Manojkumar B. Kokare , Sumit Gautam , Swaminathan R
Reconfigurable intelligent surfaces (RISs) have emerged as a highly promising technology in sixth-generation (6G) vehicular systems, offering the ability to dynamically control the wireless propagation environment. In this paper, we examine simultaneous wireless information and power transfer (SWIPT) by employing multiple RISs within a vehicle-to-infrastructure (V2I) communication system. The wireless environment exhibits high complexity due to fading and shadowing effects. To model this accurately, we adopt the double generalized Gamma (dGG) distribution. This comprehensive modeling approach enables a more realistic and insightful performance evaluation of RIS-assisted SWIPT systems under practical mobility and fading conditions. To reflect real-world vehicular dynamics, we incorporate a statistical Random Waypoint (RWP) mobility model, while also accounting for imperfections in channel state information (CSI) that arise due to high mobility and channel estimation errors. The study also integrates a non-linear energy harvesting (NL-EH) scheme to enhance performance via the power-splitting (PS) protocol. A unified objective function is proposed to jointly optimize transmit power and PS factors, aiming to maximize both the harvested energy and information rate. To address the non-convex nature of the problem, an iterative algorithm is utilized, supported by closed-form solutions derived from the Karush-Kuhn-Tucker (KKT) conditions and joint optimization (JO) method. Monte-Carlo simulations are conducted to verify the accuracy of the analytical results. Additionally, a deep neural network (DNN) framework is introduced for optimized value prediction, demonstrating superior SWIPT performance compared to single RIS configurations, with reduced complexity and faster execution.
{"title":"Optimization for dynamic multi-RIS-assisted SWIPT-Enabled V2I networks: A deep learning approach","authors":"Manojkumar B. Kokare , Sumit Gautam , Swaminathan R","doi":"10.1016/j.vehcom.2025.100984","DOIUrl":"10.1016/j.vehcom.2025.100984","url":null,"abstract":"<div><div>Reconfigurable intelligent surfaces (RISs) have emerged as a highly promising technology in sixth-generation (6G) vehicular systems, offering the ability to dynamically control the wireless propagation environment. In this paper, we examine simultaneous wireless information and power transfer (SWIPT) by employing multiple RISs within a vehicle-to-infrastructure (V2I) communication system. The wireless environment exhibits high complexity due to fading and shadowing effects. To model this accurately, we adopt the double generalized Gamma (dGG) distribution. This comprehensive modeling approach enables a more realistic and insightful performance evaluation of RIS-assisted SWIPT systems under practical mobility and fading conditions. To reflect real-world vehicular dynamics, we incorporate a statistical Random Waypoint (RWP) mobility model, while also accounting for imperfections in channel state information (CSI) that arise due to high mobility and channel estimation errors. The study also integrates a non-linear energy harvesting (NL-EH) scheme to enhance performance via the power-splitting (PS) protocol. A unified objective function is proposed to jointly optimize transmit power and PS factors, aiming to maximize both the harvested energy and information rate. To address the non-convex nature of the problem, an iterative algorithm is utilized, supported by closed-form solutions derived from the Karush-Kuhn-Tucker (KKT) conditions and joint optimization (JO) method. Monte-Carlo simulations are conducted to verify the accuracy of the analytical results. Additionally, a deep neural network (DNN) framework is introduced for optimized value prediction, demonstrating superior SWIPT performance compared to single RIS configurations, with reduced complexity and faster execution.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"57 ","pages":"Article 100984"},"PeriodicalIF":6.5,"publicationDate":"2025-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145461739","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 : 2025-11-06DOI: 10.1016/j.vehcom.2025.100987
Tong Wang
Unmanned aerial vehicle (UAV)-aided intelligent reflecting surfaces (IRSs) offer a transformative approach to enhancing wireless connectivity and coverage. This paper tackles the critical challenge of maximizing energy efficiency (EE) in such a system while guaranteeing physical layer security. We consider a network where a multi-antenna base station (BS), assisted by a UAV-mounted IRS, serves multiple ground users (GUs) in the presence of multiple eavesdroppers. To proactively secure transmissions, the BS simultaneously transmits artificial noise (AN) to degrade the eavesdroppers’ channels. We propose a holistic framework to maximize the overall system EE. Our approach orchestrates a comprehensive set of variables: the UAV’s 3D trajectory, the BS’s information and jamming beamforming, the dynamic selection of active transmit antennas at the BS, and the passive phase shifts of the IRS. This joint optimization is formulated under constraints for the GUs’ minimum secure Quality of Service (QoS), the BS’s total transmit power budget, and the UAV’s kinematic limits. The resulting problem is a highly complex, non-convex fractional program with coupled continuous and binary variables. To find a tractable solution, we design a multi-stage iterative algorithm that employs the Dinkelbach method and a Block Coordinate Descent (BCD) framework. Within each BCD iteration, the non-convex subproblems are solved using advanced techniques, including Semidefinite Relaxation (SDR), Successive Convex Approximation (SCA), and the Big-M method. Simulation results demonstrate that our orchestrated scheme significantly outperforms various benchmarks, providing crucial insights into the synergistic benefits of jointly designing active jamming and dynamic antenna selection for secure and energy-efficient aerial networks.
{"title":"Orchestrating trajectory, active jamming, and antenna selection for energy-efficient secure aerial IRS communications","authors":"Tong Wang","doi":"10.1016/j.vehcom.2025.100987","DOIUrl":"10.1016/j.vehcom.2025.100987","url":null,"abstract":"<div><div>Unmanned aerial vehicle (UAV)-aided intelligent reflecting surfaces (IRSs) offer a transformative approach to enhancing wireless connectivity and coverage. This paper tackles the critical challenge of maximizing energy efficiency (EE) in such a system while guaranteeing physical layer security. We consider a network where a multi-antenna base station (BS), assisted by a UAV-mounted IRS, serves multiple ground users (GUs) in the presence of multiple eavesdroppers. To proactively secure transmissions, the BS simultaneously transmits artificial noise (AN) to degrade the eavesdroppers’ channels. We propose a holistic framework to maximize the overall system EE. Our approach orchestrates a comprehensive set of variables: the UAV’s 3D trajectory, the BS’s information and jamming beamforming, the dynamic selection of active transmit antennas at the BS, and the passive phase shifts of the IRS. This joint optimization is formulated under constraints for the GUs’ minimum secure Quality of Service (QoS), the BS’s total transmit power budget, and the UAV’s kinematic limits. The resulting problem is a highly complex, non-convex fractional program with coupled continuous and binary variables. To find a tractable solution, we design a multi-stage iterative algorithm that employs the Dinkelbach method and a Block Coordinate Descent (BCD) framework. Within each BCD iteration, the non-convex subproblems are solved using advanced techniques, including Semidefinite Relaxation (SDR), Successive Convex Approximation (SCA), and the Big-M method. Simulation results demonstrate that our orchestrated scheme significantly outperforms various benchmarks, providing crucial insights into the synergistic benefits of jointly designing active jamming and dynamic antenna selection for secure and energy-efficient aerial networks.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"57 ","pages":"Article 100987"},"PeriodicalIF":6.5,"publicationDate":"2025-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145461740","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 : 2025-11-04DOI: 10.1016/j.vehcom.2025.100986
Yongqiang Cui, Yiyang Zhang , Di Bai, Yi Diao, Yulei Wang
Reliable self-localization of unmanned aerial vehicles (UAVs) in dense urban environments remains a major challenge due to the frequent unavailability or degradation of Global Navigation Satellite Systems (GNSS) and other radio signals. This paper presents a robust and cost-effective method for UAV self-localization by using vision and millimeter-wave (mmWave) radar data in GNSS-denied environments. The approach generates an initial dense point cloud through depth estimation and semantic segmentation, which is then geometrically refined using sparse mmWave radar point cloud. A semantic-guided clustering method is applied to the mmWave radar point cloud to remove noise and extract key structural elements such as walls, which are later fused with vision-based depth information. For positioning, image matching algorithm provides coarse localization, followed by fine registration that leverages geometric features of windows to enhance precision. Experimental results demonstrate that the proposed method can achieve self-localization accuracy within 0.4 m, while maintaining low system complexity and deployment cost, offering a practical solution for UAV self-localization in GNSS-denied urban scenarios.
{"title":"3D map and mmWave radar-based self-localization for UAVs in GNSS-denied environments","authors":"Yongqiang Cui, Yiyang Zhang , Di Bai, Yi Diao, Yulei Wang","doi":"10.1016/j.vehcom.2025.100986","DOIUrl":"10.1016/j.vehcom.2025.100986","url":null,"abstract":"<div><div>Reliable self-localization of unmanned aerial vehicles (UAVs) in dense urban environments remains a major challenge due to the frequent unavailability or degradation of Global Navigation Satellite Systems (GNSS) and other radio signals. This paper presents a robust and cost-effective method for UAV self-localization by using vision and millimeter-wave (mmWave) radar data in GNSS-denied environments. The approach generates an initial dense point cloud through depth estimation and semantic segmentation, which is then geometrically refined using sparse mmWave radar point cloud. A semantic-guided clustering method is applied to the mmWave radar point cloud to remove noise and extract key structural elements such as walls, which are later fused with vision-based depth information. For positioning, image matching algorithm provides coarse localization, followed by fine registration that leverages geometric features of windows to enhance precision. Experimental results demonstrate that the proposed method can achieve self-localization accuracy within 0.4 m, while maintaining low system complexity and deployment cost, offering a practical solution for UAV self-localization in GNSS-denied urban scenarios.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"57 ","pages":"Article 100986"},"PeriodicalIF":6.5,"publicationDate":"2025-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145442107","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 : 2025-10-24DOI: 10.1016/j.vehcom.2025.100983
Amira A. Amer , Ihab E. Talkhan , Hattan F. Abutarboush , Tawfik Ismail
Vehicle-to-Everything (V2X) communication is essential for developing fully autonomous vehicles, but it raises significant challenges due to high data rate demands and energy consumption in dense networks. This paper proposes a novel joint optimization framework integrating vehicle clustering and power allocation in sectorized 6G networks with beamforming. The framework uses a k-medoids-based clustering algorithm and a dynamic power allocation scheme to reduce interference and minimize power consumption while meeting Service Level Agreement (SLA) requirements. Our results demonstrate that the proposed framework improves SLA compliance by up to under highly dense and variable traffic conditions compared to non-clustered networks. Furthermore, dynamic power allocation reduces communication power consumption by , and Remote Radio Head (RRH) on/off switching decreases overall system power by . This approach significantly enhances network capacity and energy efficiency, making it a promising solution for sustainable V2X communications in future autonomous vehicle networks.
{"title":"Joint vehicle clustering and dynamic power allocation optimization in sectorized 6G networks for V2X communication","authors":"Amira A. Amer , Ihab E. Talkhan , Hattan F. Abutarboush , Tawfik Ismail","doi":"10.1016/j.vehcom.2025.100983","DOIUrl":"10.1016/j.vehcom.2025.100983","url":null,"abstract":"<div><div>Vehicle-to-Everything (V2X) communication is essential for developing fully autonomous vehicles, but it raises significant challenges due to high data rate demands and energy consumption in dense networks. This paper proposes a novel joint optimization framework integrating vehicle clustering and power allocation in sectorized 6G networks with beamforming. The framework uses a k-medoids-based clustering algorithm and a dynamic power allocation scheme to reduce interference and minimize power consumption while meeting Service Level Agreement (SLA) requirements. Our results demonstrate that the proposed framework improves SLA compliance by up to <span><math><mrow><mn>98.7</mn><mspace></mspace><mo>%</mo></mrow></math></span> under highly dense and variable traffic conditions compared to non-clustered networks. Furthermore, dynamic power allocation reduces communication power consumption by <span><math><mrow><mn>69</mn><mspace></mspace><mo>%</mo></mrow></math></span>, and Remote Radio Head (RRH) on/off switching decreases overall system power by <span><math><mrow><mn>3.7</mn><mspace></mspace><mo>%</mo></mrow></math></span>. This approach significantly enhances network capacity and energy efficiency, making it a promising solution for sustainable V2X communications in future autonomous vehicle networks.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"57 ","pages":"Article 100983"},"PeriodicalIF":6.5,"publicationDate":"2025-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145383757","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 : 2025-10-21DOI: 10.1016/j.vehcom.2025.100982
Xi Liu , Jun Liu
In the context of the Internet of Things, edge computing can no longer fully meet the needs of the diversified task requirements; hence, vehicle computing has been developed. Concerning the new feature of intelligent vehicles (IVs) mobility, a novel elastic mobile service model is proposed for vehicle computing, where vehicle computing dynamically changes computing power according to the demands. This paper considers the elastic mobile service in two communication scenarios. The first is that IVs from different areas quickly move to the area with insufficient computing power, and the second is that IVs move to the specified area by the user. Owing to the heterogeneous costs of IVs, the pricing is based on the auction. Our goal is to maximize the total value, which is the sum of the values of all tasks running on the IVs. The problem is formulated as an integer linear program and proven -hard. Then, a budget-feasible strategy-proof mechanism is proposed to allocate IVs one by one based on the marginal contribution. We first show that the proposed mechanism achieves strategy-proofness, individual rationality, consumer sovereignty, and budget balance, which drives the system into equilibrium. The approximation ratio of the proposed mechanism is analyzed. The experimental results show that under budget constraints, the proposed mechanism not only approaches the optimal performance in terms of total value but also effectively allocates tasks to provide the elastic mobile service.
{"title":"Budget feasible strategy-proof mechanism design for elastic mobile service in vehicle computing","authors":"Xi Liu , Jun Liu","doi":"10.1016/j.vehcom.2025.100982","DOIUrl":"10.1016/j.vehcom.2025.100982","url":null,"abstract":"<div><div>In the context of the Internet of Things, edge computing can no longer fully meet the needs of the diversified task requirements; hence, vehicle computing has been developed. Concerning the new feature of intelligent vehicles (IVs) mobility, a novel elastic mobile service model is proposed for vehicle computing, where vehicle computing dynamically changes computing power according to the demands. This paper considers the elastic mobile service in two communication scenarios. The first is that IVs from different areas quickly move to the area with insufficient computing power, and the second is that IVs move to the specified area by the user. Owing to the heterogeneous costs of IVs, the pricing is based on the auction. Our goal is to maximize the total value, which is the sum of the values of all tasks running on the IVs. The problem is formulated as an integer linear program and proven <span><math><mrow><mi>N</mi><mi>P</mi></mrow></math></span>-hard. Then, a budget-feasible strategy-proof mechanism is proposed to allocate IVs one by one based on the marginal contribution. We first show that the proposed mechanism achieves strategy-proofness, individual rationality, consumer sovereignty, and budget balance, which drives the system into equilibrium. The approximation ratio of the proposed mechanism is analyzed. The experimental results show that under budget constraints, the proposed mechanism not only approaches the optimal performance in terms of total value but also effectively allocates tasks to provide the elastic mobile service.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"56 ","pages":"Article 100982"},"PeriodicalIF":6.5,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145397951","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 : 2025-10-15DOI: 10.1016/j.vehcom.2025.100981
Peiying Zhang , Renzhuang Yuan , Lizhuang Tan , Konstantin Igorevich Kostromitin , Athanasios V. Vasilakos , Jian Wang
With the rapid development of Intelligent Vehicles and Mobile Edge Computing, the Internet of Vehicles (IoV) faces numerous challenges when handling computationally intensive tasks, primarily including limited computing resources and dynamic requirement distributions in the IoV system. To address these challenges, this paper proposes a novel edge–cloud collaborative offloading framework for the Internet of Vehicles, named TP-IoV. This framework achieves proactive optimization of computation task offloading by combining time series task prediction with an adaptive decision mechanism. Specifically, TP-IoV utilizes the Informer model to predict future task characteristics and employs the Twin Delayed Deep Deterministic Policy Gradient algorithm to determine the optimal allocation strategy for tasks among local vehicles, edge nodes, and cloud servers. Simulation results demonstrate that under high load and strict latency conditions, TP-IoV significantly outperforms existing baseline methods in reducing task latency and improving the task completion success rate. This work provides a scalable and proactive solution for latency-sensitive applications in dynamic Internet of Vehicles environments.
随着智能汽车和移动边缘计算的快速发展，车联网在处理计算密集型任务时面临着许多挑战，主要包括车联网系统中有限的计算资源和动态的需求分布。为了应对这些挑战，本文提出了一种新型的车联网边缘云协作卸载框架，称为TP-IoV。该框架将时间序列任务预测与自适应决策机制相结合，实现了计算任务卸载的主动优化。具体而言，TP-IoV利用Informer模型预测未来任务特征，并采用Twin Delayed Deep Deterministic Policy Gradient算法确定任务在本地车辆、边缘节点和云服务器之间的最优分配策略。仿真结果表明，在高负载和严格延迟条件下，TP-IoV在降低任务延迟和提高任务完成成功率方面明显优于现有的基线方法。这项工作为动态车联网环境中对延迟敏感的应用提供了可扩展的主动解决方案。
{"title":"TP-IoV: A task prediction-oriented cloud-edge collaborative offloading framework for Internet of vehicles","authors":"Peiying Zhang , Renzhuang Yuan , Lizhuang Tan , Konstantin Igorevich Kostromitin , Athanasios V. Vasilakos , Jian Wang","doi":"10.1016/j.vehcom.2025.100981","DOIUrl":"10.1016/j.vehcom.2025.100981","url":null,"abstract":"<div><div>With the rapid development of Intelligent Vehicles and Mobile Edge Computing, the Internet of Vehicles (IoV) faces numerous challenges when handling computationally intensive tasks, primarily including limited computing resources and dynamic requirement distributions in the IoV system. To address these challenges, this paper proposes a novel edge–cloud collaborative offloading framework for the Internet of Vehicles, named TP-IoV. This framework achieves proactive optimization of computation task offloading by combining time series task prediction with an adaptive decision mechanism. Specifically, TP-IoV utilizes the Informer model to predict future task characteristics and employs the Twin Delayed Deep Deterministic Policy Gradient algorithm to determine the optimal allocation strategy for tasks among local vehicles, edge nodes, and cloud servers. Simulation results demonstrate that under high load and strict latency conditions, TP-IoV significantly outperforms existing baseline methods in reducing task latency and improving the task completion success rate. This work provides a scalable and proactive solution for latency-sensitive applications in dynamic Internet of Vehicles environments.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"56 ","pages":"Article 100981"},"PeriodicalIF":6.5,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145363500","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 : 2025-10-10DOI: 10.1016/j.vehcom.2025.100980
Aiswarya S. Nair , Sabu M. Thampi , Jithu Vijay V. P.
GPS spoofing remains a significant and persistent threat to Internet of Drones (IoD), which compromises navigation integrity, security, and reliability. Drones, constrained by limited computational resources and power, demand innovative solutions to combat this easily exploitable vulnerability. Existing detection methods lack computational efficiency, contextual intelligence, and collaborative validation, leading to high false positives and low adaptability. In this paper, we propose a context-aware GPS spoofing detection and mitigation framework, SoCoMNNet, that integrates Memristive Neural Networks (MNNs) and a SocioCognitive fuzzy inference system for trust-driven behaviour analysis. The MNN module, deployed on each drone, detects navigation inconsistencies with minimal computational overhead, while the SocioCognitive system at the Ground Control Station (GCS) evaluates drone's behaviour in terms of Ability, Benevolence, and Integrity (ABI) to differentiate adversarial GPS spoofing from mission deviations. The predictions from the MNN and the behaviour assessment are combined using a weighted average, where both are given equal importance. In this way, the final result considers what the model predicts as well as how the drone is actually behaving, making GPS spoofing detection more accurate and context-aware. The contextual understanding provided by the SocioCognitive fuzzy system helps differentiate intentional deviations from unexpected ones, enhancing the overall resilience of the system. We have also developed a Kyber Post-Quantum Cryptography (PQC) secured GPS spoofing mitigation mechanism that helps drones to recover authentic GPS data during spoofing attacks. We evaluated the performance of MNN using MemTorch for memristor-based neural modelling, and NeuroSIM for hardware-level simulation and resource analysis. The fuzzy inference engine runs 27 rules and deduces five drone behaviours such as Discard, Unsatisfactory, Satisfactory, Reliable, and Ideal. Incorporating this context awareness into the detection process enables SoCoMNNet to reduce false positives during GPS spoofing detection. A statistical t-test was performed to show the impact of the proposed detection approach. The Kyber PQC mitigation approach was evaluated on Raspberry Pi 4 in terms of computation cost, communication overhead, and storage requirements. The results show reduced execution time, higher computational efficiency, lower memory usage, and stronger system security. Our integrated solution delivers a resilient and computationally efficient security framework for IoD in adversarial GPS spoofing environments.
GPS欺骗仍然是无人机互联网（IoD）的一个重大而持久的威胁，它损害了导航的完整性、安全性和可靠性。无人机受到有限的计算资源和能力的限制，需要创新的解决方案来对抗这种容易被利用的漏洞。现有检测方法缺乏计算效率、上下文智能和协同验证，导致误报率高、适应性低。在本文中，我们提出了一个上下文感知的GPS欺骗检测和缓解框架SoCoMNNet，它集成了记忆神经网络（MNNs）和社会认知模糊推理系统，用于信任驱动的行为分析。部署在每架无人机上的MNN模块，以最小的计算开销检测导航不一致，而地面控制站（GCS）的社会认知系统（social cognitive system）根据能力、仁慈和完整性（ABI）评估无人机的行为，以区分对抗性GPS欺骗和任务偏差。来自MNN的预测和行为评估使用加权平均值进行组合，其中两者具有同等的重要性。通过这种方式，最终结果考虑了模型预测的内容以及无人机的实际行为，使GPS欺骗检测更加准确和具有上下文感知能力。社会认知模糊系统提供的上下文理解有助于区分有意偏差和意外偏差，增强系统的整体弹性。我们还开发了Kyber后量子加密（PQC）安全的GPS欺骗缓解机制，帮助无人机在欺骗攻击期间恢复真实的GPS数据。我们使用MemTorch进行基于忆阻器的神经建模，使用NeuroSIM进行硬件级仿真和资源分析，评估了MNN的性能。模糊推理引擎运行27条规则，演绎出丢弃、不满意、满意、可靠、理想五种无人机行为。将这种上下文感知整合到检测过程中，SoCoMNNet可以减少GPS欺骗检测过程中的误报。进行统计t检验以显示所提出的检测方法的影响。Kyber PQC缓解方法在Raspberry Pi 4上进行了计算成本、通信开销和存储需求方面的评估。结果表明，减少了执行时间，提高了计算效率，降低了内存使用，增强了系统安全性。我们的集成解决方案为对抗GPS欺骗环境中的IoD提供了弹性和计算效率高的安全框架。
{"title":"SoCoMNNet: A SocioCognitive and memristive neural network-based context-aware GPS spoofing detection and mitigation in the Internet of drones","authors":"Aiswarya S. Nair , Sabu M. Thampi , Jithu Vijay V. P.","doi":"10.1016/j.vehcom.2025.100980","DOIUrl":"10.1016/j.vehcom.2025.100980","url":null,"abstract":"<div><div>GPS spoofing remains a significant and persistent threat to Internet of Drones (IoD), which compromises navigation integrity, security, and reliability. Drones, constrained by limited computational resources and power, demand innovative solutions to combat this easily exploitable vulnerability. Existing detection methods lack computational efficiency, contextual intelligence, and collaborative validation, leading to high false positives and low adaptability. In this paper, we propose a context-aware GPS spoofing detection and mitigation framework, SoCoMNNet, that integrates Memristive Neural Networks (MNNs) and a SocioCognitive fuzzy inference system for trust-driven behaviour analysis. The MNN module, deployed on each drone, detects navigation inconsistencies with minimal computational overhead, while the SocioCognitive system at the Ground Control Station (GCS) evaluates drone's behaviour in terms of Ability, Benevolence, and Integrity (ABI) to differentiate adversarial GPS spoofing from mission deviations. The predictions from the MNN and the behaviour assessment are combined using a weighted average, where both are given equal importance. In this way, the final result considers what the model predicts as well as how the drone is actually behaving, making GPS spoofing detection more accurate and context-aware. The contextual understanding provided by the SocioCognitive fuzzy system helps differentiate intentional deviations from unexpected ones, enhancing the overall resilience of the system. We have also developed a Kyber Post-Quantum Cryptography (PQC) secured GPS spoofing mitigation mechanism that helps drones to recover authentic GPS data during spoofing attacks. We evaluated the performance of MNN using MemTorch for memristor-based neural modelling, and NeuroSIM for hardware-level simulation and resource analysis. The fuzzy inference engine runs 27 rules and deduces five drone behaviours such as Discard, Unsatisfactory, Satisfactory, Reliable, and Ideal. Incorporating this context awareness into the detection process enables SoCoMNNet to reduce false positives during GPS spoofing detection. A statistical t-test was performed to show the impact of the proposed detection approach. The Kyber PQC mitigation approach was evaluated on Raspberry Pi 4 in terms of computation cost, communication overhead, and storage requirements. The results show reduced execution time, higher computational efficiency, lower memory usage, and stronger system security. Our integrated solution delivers a resilient and computationally efficient security framework for IoD in adversarial GPS spoofing environments.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"56 ","pages":"Article 100980"},"PeriodicalIF":6.5,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145321171","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 : 2025-10-10DOI: 10.1016/j.vehcom.2025.100979
Youqiang Hu
Vehicular federated learning is able to deal with the data shortage dilemma in practical Artificial Intelligence (AI) projects in Internet of Vehicles (IoV) scenarios, but there is a bottleneck in this training paradigm, that is, the energy consumption issue. Since the training process is carried out on vehicles and the training tasks are generally computation-intensive, the battery lives of vehicles will be greatly affected if they participate in training. To improve the sustainability of the participants, split learning is introduced to the procedure of vehicular federated learning in this paper. Split learning divides the trained AI model into two parts. One is retained on terminals, and the other is sent to the cloud servers for remote training. Since the computation workload is reduced, the energy consumptions of participants are lowered. This training paradigm is called Vehicular Federated Split Learning (VFSL). Then, we model the computation and communication processes of VFSL and derive the energy consumption minimization problem. The estimated Channel State Information (CSI) between high-moving vehicles and Road Side Units (RSUs) is generally inaccurate. The imperfect CSI makes the formulated problem a stochastic mixed integer nonlinear programming problem, which is hard to solve. Therefore, we propose a resource allocation and model split strategy based on the Constrained Stochastic Successive Convex Approximation (CSSCA) and greedy algorithms. Simulation results demonstrate that the proposed strategy is able to achieve higher energy efficiency compared to the existing strategies in the case of imperfect CSI.
{"title":"Resource allocation and model split for energy-efficient federated split learning in Internet of Vehicles with imperfect CSI","authors":"Youqiang Hu","doi":"10.1016/j.vehcom.2025.100979","DOIUrl":"10.1016/j.vehcom.2025.100979","url":null,"abstract":"<div><div>Vehicular federated learning is able to deal with the data shortage dilemma in practical Artificial Intelligence (AI) projects in Internet of Vehicles (IoV) scenarios, but there is a bottleneck in this training paradigm, that is, the energy consumption issue. Since the training process is carried out on vehicles and the training tasks are generally computation-intensive, the battery lives of vehicles will be greatly affected if they participate in training. To improve the sustainability of the participants, split learning is introduced to the procedure of vehicular federated learning in this paper. Split learning divides the trained AI model into two parts. One is retained on terminals, and the other is sent to the cloud servers for remote training. Since the computation workload is reduced, the energy consumptions of participants are lowered. This training paradigm is called Vehicular Federated Split Learning (VFSL). Then, we model the computation and communication processes of VFSL and derive the energy consumption minimization problem. The estimated Channel State Information (CSI) between high-moving vehicles and Road Side Units (RSUs) is generally inaccurate. The imperfect CSI makes the formulated problem a stochastic mixed integer nonlinear programming problem, which is hard to solve. Therefore, we propose a resource allocation and model split strategy based on the Constrained Stochastic Successive Convex Approximation (CSSCA) and greedy algorithms. Simulation results demonstrate that the proposed strategy is able to achieve higher energy efficiency compared to the existing strategies in the case of imperfect CSI.</div></div>","PeriodicalId":54346,"journal":{"name":"Vehicular Communications","volume":"56 ","pages":"Article 100979"},"PeriodicalIF":6.5,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145321169","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