Pub Date : 2025-12-03DOI: 10.1109/OJVT.2025.3639480
Chao Song;Hao Li;Liangliang Huai;Shuangshuang Luo;Bo Li;Kaifang Wan
To address the challenges of limited autonomy, low decision-making efficiency, and poor generalization in UAV task planning for tracking mobile target under uncertain situations, this paper proposes a transfer-fusion algorithm based on the integration of three-way decision-making and self-attention mechanism into an optimized Soft Actor-Critic framework (TW-AM-SAC). Unlike research that mostly turns to deterministic reinforcement learning strategy, this one introduces a non-deterministic SAC algorithm to integrate the exploration and improvement into a single strategy to help realize the UAV’s autonomous decision-making. Subsequently, to mitigate the issues of singular reward functions with fixed weights in task planning, three-way decision-making theory is incorporated to design autonomous reward functions tailored to different situations, while a self-attention mechanism is fused to assign dynamic weight distributions to the reward components. Furthermore, to enhance the adaptability of the intelligent algorithm across varying situations, a transfer learning model incorporating self- game is constructed to improve generalization performance. The simulation verification can be known that the TW-AM-SAC transfer-algorithm proposed in this paper has more effective tracking frequency and greater advantages in autonomous tracking when applied to UAV tracking of moving targets, and meanwhile converges faster with better generalization, compared with the single SAC algorithm.
{"title":"UAV’s Task Planning for Tracking the Moving Target Based on TW-AM-SAC Transfer Fusion Algorithm","authors":"Chao Song;Hao Li;Liangliang Huai;Shuangshuang Luo;Bo Li;Kaifang Wan","doi":"10.1109/OJVT.2025.3639480","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3639480","url":null,"abstract":"To address the challenges of limited autonomy, low decision-making efficiency, and poor generalization in UAV task planning for tracking mobile target under uncertain situations, this paper proposes a transfer-fusion algorithm based on the integration of three-way decision-making and self-attention mechanism into an optimized Soft Actor-Critic framework (TW-AM-SAC). Unlike research that mostly turns to deterministic reinforcement learning strategy, this one introduces a non-deterministic SAC algorithm to integrate the exploration and improvement into a single strategy to help realize the UAV’s autonomous decision-making. Subsequently, to mitigate the issues of singular reward functions with fixed weights in task planning, three-way decision-making theory is incorporated to design autonomous reward functions tailored to different situations, while a self-attention mechanism is fused to assign dynamic weight distributions to the reward components. Furthermore, to enhance the adaptability of the intelligent algorithm across varying situations, a transfer learning model incorporating self- game is constructed to improve generalization performance. The simulation verification can be known that the TW-AM-SAC transfer-algorithm proposed in this paper has more effective tracking frequency and greater advantages in autonomous tracking when applied to UAV tracking of moving targets, and meanwhile converges faster with better generalization, compared with the single SAC algorithm.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"273-289"},"PeriodicalIF":4.8,"publicationDate":"2025-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11275675","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145929683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-28DOI: 10.1109/OJVT.2025.3638680
Reza Jafari;Shady S. Refaat;Amin Paykani;Pedram Asef;Pouria Sarhadi
Torque vectoring can enhance dynamic stability and concurrently enable efficient energy management in electric vehicles (EVs) through optimized torque distribution. Nevertheless, conventional torque vectoring schemes often rely on fixed models and tuning, limiting their adaptability. Reinforcement learning (RL) and its model-free versions employing deep neural networks allow the development of control policies through direct interaction with the environment, making it suitable for complex and nonlinear dynamics. This paper presents a comprehensive survey of recent research on the application of RL for torque vectoring and energy optimization in EVs. An overview of conventional direct yaw control (DYC) approaches, their objectives, and common hierarchical strategies are initially studied to establish a foundation for discussing model-free RL-based torque vectoring. A description of RL in the context of stability-oriented control and energy optimization, key components, operational processes, and their classifications are studied. The primary emphasis is on RL-based torque vectoring and energy management in EVs to improve yaw stability, reduce energy consumption, and manage trade-offs under real-time constraints. Overall, RL-based controllers provide enhanced adaptability to modeling inaccuracies and facilitate more straightforward multi-objective design for simultaneous energy management and stability control, making them promising alternatives to conventional model-based methods.
{"title":"Reinforcement Learning for Torque Vectoring in Electric Vehicles: A Review of Stability and Energy Optimization Methods","authors":"Reza Jafari;Shady S. Refaat;Amin Paykani;Pedram Asef;Pouria Sarhadi","doi":"10.1109/OJVT.2025.3638680","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3638680","url":null,"abstract":"Torque vectoring can enhance dynamic stability and concurrently enable efficient energy management in electric vehicles (EVs) through optimized torque distribution. Nevertheless, conventional torque vectoring schemes often rely on fixed models and tuning, limiting their adaptability. Reinforcement learning (RL) and its model-free versions employing deep neural networks allow the development of control policies through direct interaction with the environment, making it suitable for complex and nonlinear dynamics. This paper presents a comprehensive survey of recent research on the application of RL for torque vectoring and energy optimization in EVs. An overview of conventional direct yaw control (DYC) approaches, their objectives, and common hierarchical strategies are initially studied to establish a foundation for discussing model-free RL-based torque vectoring. A description of RL in the context of stability-oriented control and energy optimization, key components, operational processes, and their classifications are studied. The primary emphasis is on RL-based torque vectoring and energy management in EVs to improve yaw stability, reduce energy consumption, and manage trade-offs under real-time constraints. Overall, RL-based controllers provide enhanced adaptability to modeling inaccuracies and facilitate more straightforward multi-objective design for simultaneous energy management and stability control, making them promising alternatives to conventional model-based methods.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"354-380"},"PeriodicalIF":4.8,"publicationDate":"2025-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11271302","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145982285","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-28DOI: 10.1109/OJVT.2025.3638462
Pramod N. Chine;Suven Jagtiani;Mandar R. Nalavade;Gaurav S. Kasbekar
In wireless networks, algorithms for user association, i.e., the task of choosing the base station (BS) that every arriving user should join, significantly impact the network performance. A wireless network with multiple BSs, operating on non-overlapping channels, is considered. The channels of the BSs are susceptible to jamming by attackers. During every time slot, a user arrives with a certain probability. There exists a holding cost in each slot for every user associated with a BS. The goal here is to design a user association scheme, which assigns a BS to each user upon arrival, with the objective of minimizing the long-run total average holding cost borne within the network. This objective results in low average delays attained by users. This association problem is an instance of restless multi-armed bandit problems, and is known to be hard to solve. By making use of the framework presented by Whittle, the hard per-stage constraint that every arriving user must connect to exactly one BS in a time slot is relaxed to a long-term time-averaged constraint. Subsequently, we employ the Lagrangian multiplier strategy to reformulate the problem into an unconstrained form and decompose it into separate Markov decision processes at the BSs. Further, the problem is proven to be Whittle indexable and a method for calculating the Whittle indices corresponding to different BSs is presented. We design a user association policy under which, upon arrival of a user in a time slot, it is assigned to the BS having the least Whittle index in that slot. This research is significant as it provides a scalable and resilient decision-making framework for user association in adversarial wireless environments. The proposed Whittle index-based policy achieves low long-term expected average cost, robustness to jamming, and improved average delay and fairness performance. However, its effectiveness depends on accurate estimation of system parameters and may be limited under highly dynamic network conditions. Through extensive simulations, we show that our proposed association policy outperforms various user association policies proposed in previous work in terms of different metrics such as average cost, average delay, and Jain’s fairness index.
{"title":"User Association in the Presence of Jamming in Wireless Networks Using the Whittle Index","authors":"Pramod N. Chine;Suven Jagtiani;Mandar R. Nalavade;Gaurav S. Kasbekar","doi":"10.1109/OJVT.2025.3638462","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3638462","url":null,"abstract":"In wireless networks, algorithms for user association, i.e., the task of choosing the base station (BS) that every arriving user should join, significantly impact the network performance. A wireless network with multiple BSs, operating on non-overlapping channels, is considered. The channels of the BSs are susceptible to jamming by attackers. During every time slot, a user arrives with a certain probability. There exists a holding cost in each slot for every user associated with a BS. The goal here is to design a user association scheme, which assigns a BS to each user upon arrival, with the objective of minimizing the long-run total average holding cost borne within the network. This objective results in low average delays attained by users. This association problem is an instance of restless multi-armed bandit problems, and is known to be hard to solve. By making use of the framework presented by Whittle, the hard per-stage constraint that every arriving user must connect to exactly one BS in a time slot is relaxed to a long-term time-averaged constraint. Subsequently, we employ the Lagrangian multiplier strategy to reformulate the problem into an unconstrained form and decompose it into separate Markov decision processes at the BSs. Further, the problem is proven to be Whittle indexable and a method for calculating the Whittle indices corresponding to different BSs is presented. We design a user association policy under which, upon arrival of a user in a time slot, it is assigned to the BS having the least Whittle index in that slot. This research is significant as it provides a scalable and resilient decision-making framework for user association in adversarial wireless environments. The proposed Whittle index-based policy achieves low long-term expected average cost, robustness to jamming, and improved average delay and fairness performance. However, its effectiveness depends on accurate estimation of system parameters and may be limited under highly dynamic network conditions. Through extensive simulations, we show that our proposed association policy outperforms various user association policies proposed in previous work in terms of different metrics such as average cost, average delay, and Jain’s fairness index.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"194-214"},"PeriodicalIF":4.8,"publicationDate":"2025-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11271329","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145778472","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This tutorial describes the 5G New Radio Vehicle-to-Everything (5G NR-V2X) air interface, with a specific focus on the features and capabilities introduced in 3GPP Release 16. It begins by outlining the motivation for 5G NR-V2X and then progresses to the standardized definitions of the air interface, upper layer standards, and application protocols. Simulated performance on two classes of applications, urban intersection and highway merge is presented, leading to a conclusion that the lower layer standardization can address maneuver coordination – where nearby vehicles could effectively communicate to and therefore cooperate with nearby relevant vehicles. This portends a next and perhaps concluding step in realizing the full benefits of Cooperative, Connected, and Automated Mobility (CCAM) in Europe and down the line, in other global regions.
{"title":"A Tutorial on 5G NR-V2X: Enhancements, Real-World Applications, and Performance Evaluation","authors":"Abolfazl Hajisami;Ralf Weber;Jim Misener;Ahmed Farhan Hanif","doi":"10.1109/OJVT.2025.3637712","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3637712","url":null,"abstract":"This tutorial describes the 5G New Radio Vehicle-to-Everything (5G NR-V2X) air interface, with a specific focus on the features and capabilities introduced in 3GPP Release 16. It begins by outlining the motivation for 5G NR-V2X and then progresses to the standardized definitions of the air interface, upper layer standards, and application protocols. Simulated performance on two classes of applications, urban intersection and highway merge is presented, leading to a conclusion that the lower layer standardization can address maneuver coordination – where nearby vehicles could effectively communicate to and therefore cooperate with nearby relevant vehicles. This portends a next and perhaps concluding step in realizing the full benefits of Cooperative, Connected, and Automated Mobility (CCAM) in Europe and down the line, in other global regions.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"41-53"},"PeriodicalIF":4.8,"publicationDate":"2025-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11269721","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145729520","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-24DOI: 10.1109/OJVT.2025.3636075
Imad Ali Shah;Jiarong Li;Roshan George;Tim Brophy;Enda Ward;Martin Glavin;Edward Jones;Brian Deegan
Hyperspectral imaging (HSI) is a transformative sensing modality for Advanced Driver Assistance Systems (ADAS) and autonomous driving (AD). By capturing fine spectral resolution across hundreds of bands, HSI enables material-level scene understanding that overcomes critical limitations of traditional RGB imaging in adverse weather and lighting. This paper presents the first comprehensive review of HSI for automotive applications, examining the strengths, limitations, and suitability of current HSI technologies in the context of ADAS/AD. In addition, we analyze 216 commercially available spectral imaging cameras, benchmarking them against key automotive criteria: frame rate, spatial resolution, spectral dimensionality, and compliance with AEC-Q100 temperature standards. Our analysis reveals a significant gap between HSI’s demonstrated research potential and its commercial readiness. Only four cameras meet the defined performance thresholds, and none comply with AEC-Q100 requirements. In addition, the paper reviews recent HSI datasets and applications, including semantic segmentation for road surface classification, pedestrian separability, and adverse weather perception. Our review shows that current HSI datasets are limited in scale, spectral consistency, channel count, and environmental diversity, posing a challenge for perception algorithms development and adequate HSI’s potential validation in ADAS/AD applications. This review paper presents the current state of HSI in automotive contexts and outlines key research directions toward practical integration of spectral imaging in ADAS and autonomous systems.
{"title":"Hyperspectral Sensors and Autonomous Driving: Technologies, Limitations, and Opportunities","authors":"Imad Ali Shah;Jiarong Li;Roshan George;Tim Brophy;Enda Ward;Martin Glavin;Edward Jones;Brian Deegan","doi":"10.1109/OJVT.2025.3636075","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3636075","url":null,"abstract":"Hyperspectral imaging (HSI) is a transformative sensing modality for Advanced Driver Assistance Systems (ADAS) and autonomous driving (AD). By capturing fine spectral resolution across hundreds of bands, HSI enables material-level scene understanding that overcomes critical limitations of traditional RGB imaging in adverse weather and lighting. This paper presents the first comprehensive review of HSI for automotive applications, examining the strengths, limitations, and suitability of current HSI technologies in the context of ADAS/AD. In addition, we analyze 216 commercially available spectral imaging cameras, benchmarking them against key automotive criteria: frame rate, spatial resolution, spectral dimensionality, and compliance with AEC-Q100 temperature standards. Our analysis reveals a significant gap between HSI’s demonstrated research potential and its commercial readiness. Only four cameras meet the defined performance thresholds, and none comply with AEC-Q100 requirements. In addition, the paper reviews recent HSI datasets and applications, including semantic segmentation for road surface classification, pedestrian separability, and adverse weather perception. Our review shows that current HSI datasets are limited in scale, spectral consistency, channel count, and environmental diversity, posing a challenge for perception algorithms development and adequate HSI’s potential validation in ADAS/AD applications. This review paper presents the current state of HSI in automotive contexts and outlines key research directions toward practical integration of spectral imaging in ADAS and autonomous systems.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"124-143"},"PeriodicalIF":4.8,"publicationDate":"2025-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11266884","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145778470","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-20DOI: 10.1109/OJVT.2025.3635022
Javier Santiago Olmos Medina;Jessica Gissella Maradey Lázaro;Anton Rassõlkin;Mahmoud Ibrahim
The perception systems for Traffic Sign Recognition (TSR) and Lane Line Recognition (LLR) are foundational pillars for the safe and effective operation of Advanced Driver-Assistance Systems (ADAS) and fully autonomous vehicles. This review provides a comprehensive analysis of the latest academic research in these domains, strictly focusing on literature published from October 2024 to the present. The analysis reveals several key trends shaping the field. In TSR, architectural evolution is characterized by the refinement of Convolutional Neural Networks (CNNs), the specialization of light-weight YOLO-based models for real-time embedded applications, and the emergence of hybrid CNN-Transformer architectures. Concurrently, a significant research thrust is dedicated to enhancing robustness against environmental adversities and a growing spectrum of sophisticated, physically plausible adversarial attacks. In LLR, the paradigm is rapidly shifting from 2D image-plane detection to full 3D spatial localization and topology reasoning, driven by Transformer-based models that excel at capturing global context and long-range dependencies. Cross-cutting themes common to both domains include a relentless drive for computational efficiency, a data-centric approach marked by the creation of new, challenging benchmarks for adverse conditions and 3D perception, and the nascent but transformative integration of multi-task learning and Vision-Language Models (VLMs) to build systems capable of holistic scene reasoning. Despite significant progress, several key challenges persist in the field of domain generalization, particularly in handling long-tail corner cases and developing safety-aware evaluation metrics. Future research is expected to focus on self-supervised learning, stronger integration between perception and control systems, and the advancement of trustworthy AI through improved explainability and robust-ness. These efforts will lay the groundwork for the next generation of intelligent vehicle systems.
{"title":"The Road Ahead: A Comprehensive Review of Recent Advances in Traffic Sign and Lane Line Recognition for Autonomous Systems","authors":"Javier Santiago Olmos Medina;Jessica Gissella Maradey Lázaro;Anton Rassõlkin;Mahmoud Ibrahim","doi":"10.1109/OJVT.2025.3635022","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3635022","url":null,"abstract":"The perception systems for Traffic Sign Recognition (TSR) and Lane Line Recognition (LLR) are foundational pillars for the safe and effective operation of Advanced Driver-Assistance Systems (ADAS) and fully autonomous vehicles. This review provides a comprehensive analysis of the latest academic research in these domains, strictly focusing on literature published from October 2024 to the present. The analysis reveals several key trends shaping the field. In TSR, architectural evolution is characterized by the refinement of Convolutional Neural Networks (CNNs), the specialization of light-weight YOLO-based models for real-time embedded applications, and the emergence of hybrid CNN-Transformer architectures. Concurrently, a significant research thrust is dedicated to enhancing robustness against environmental adversities and a growing spectrum of sophisticated, physically plausible adversarial attacks. In LLR, the paradigm is rapidly shifting from 2D image-plane detection to full 3D spatial localization and topology reasoning, driven by Transformer-based models that excel at capturing global context and long-range dependencies. Cross-cutting themes common to both domains include a relentless drive for computational efficiency, a data-centric approach marked by the creation of new, challenging benchmarks for adverse conditions and 3D perception, and the nascent but transformative integration of multi-task learning and Vision-Language Models (VLMs) to build systems capable of holistic scene reasoning. Despite significant progress, several key challenges persist in the field of domain generalization, particularly in handling long-tail corner cases and developing safety-aware evaluation metrics. Future research is expected to focus on self-supervised learning, stronger integration between perception and control systems, and the advancement of trustworthy AI through improved explainability and robust-ness. These efforts will lay the groundwork for the next generation of intelligent vehicle systems.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"160-178"},"PeriodicalIF":4.8,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11261861","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145778471","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Lithium-ion batteries are pivotal to the energy transition, powering electric vehicles and enabling stationary energy storage systems. However, their reliance on finite and scarce materials underscores the need for improved sustainability. Extending battery lifetime by mitigating degradation mechanisms is therefore essential to enhance performance, reduce resource dependency, and support large-scale energy deployment. That requires a thorough understanding of the factors that accelerate battery aging and the strategies to optimize their usage. To that end, we propose a model for predicting the state of health of the lithium-ion battery based on a combination of convolution, Transformers and Bi-LSTM (Long Short-Term Memory), which involves using explainability methods in order to understand the inner workings and reasoning of the model. That approach predicts the capacity degradation curve from a sliding window of time series, each made up of 3 charge and discharge cycles of our laboratory dataset including current, voltage, temperature and state of charge (SOC). An extension of Shapley values for time series adapted to the problem of battery aging is proposed, allowing the study of the influence of the model input parameters from multiple perspectives, including state of charge, temperature dynamics, and current regimes. Those Shapley values quantify the influence of individual features on the battery aging rate, thereby enabling the identification of usage patterns that contribute to accelerated degradation.
{"title":"Time Series-Based Explainable Model for Lithium-Ion Battery State of Health Prediction","authors":"Théo Heitzmann;Tedjani Mesbahi;Ahmed Samet;Romuald Boné","doi":"10.1109/OJVT.2025.3634139","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3634139","url":null,"abstract":"Lithium-ion batteries are pivotal to the energy transition, powering electric vehicles and enabling stationary energy storage systems. However, their reliance on finite and scarce materials underscores the need for improved sustainability. Extending battery lifetime by mitigating degradation mechanisms is therefore essential to enhance performance, reduce resource dependency, and support large-scale energy deployment. That requires a thorough understanding of the factors that accelerate battery aging and the strategies to optimize their usage. To that end, we propose a model for predicting the state of health of the lithium-ion battery based on a combination of convolution, Transformers and Bi-LSTM (Long Short-Term Memory), which involves using explainability methods in order to understand the inner workings and reasoning of the model. That approach predicts the capacity degradation curve from a sliding window of time series, each made up of 3 charge and discharge cycles of our laboratory dataset including current, voltage, temperature and state of charge (SOC). An extension of Shapley values for time series adapted to the problem of battery aging is proposed, allowing the study of the influence of the model input parameters from multiple perspectives, including state of charge, temperature dynamics, and current regimes. Those Shapley values quantify the influence of individual features on the battery aging rate, thereby enabling the identification of usage patterns that contribute to accelerated degradation.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"144-159"},"PeriodicalIF":4.8,"publicationDate":"2025-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11251150","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145778433","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-18DOI: 10.1109/OJVT.2025.3634375
Liqiang Wang;Meng Wang
The increasing complexity of perception and decision-making tasks in intelligent connected vehicles has driven the evolution of on-board computing platforms toward heterogeneous architectures. However, the dynamic nature of workloads, the need for multi-objective optimization, and stringent safety constraints pose significant challenges to scheduling. To address the limitations of existing approaches in balancing multiple objectives and ensuring safety, this paper proposes a deep reinforcement learning (DRL)-based hierarchical hybrid-action multi-objective adaptive scheduling framework. The framework optimizes latency, energy consumption, reliability, and thermal management by introducing a dynamic weight adjustment mechanism driven by the battery state of charge (SOC) and thermal accumulation. It integrates high-level global task allocation with low-level real-time resource adjustment for adaptive multi-objective trade-offs, while embedding a functional safety fallback mechanism to guarantee hard real-time performance and thermal safety for high-criticality tasks. Experimental results under highway cruising, urban congestion, and high-temperature scenarios show that the proposed method outperforms HEFT, E-List, and Vanilla-DRL in p95 latency, energy consumption, peak temperature, and high-criticality task satisfaction: p95 latency is reduced by 6%–14%, energy consumption by 5%–20%, peak temperature by 2–8°C, and satisfaction rates exceed 97.5%. After model compression, the strategy network achieves inference latency under 5 ms and nearly 40% power reduction on an automotive-grade heterogeneous platform, validating the engineering feasibility of the approach. This work provides a scalable and safety-aware solution for efficient heterogeneous computing scheduling in intelligent vehicles.
{"title":"Deep Reinforcement Learning-Based Adaptive Scheduling for Intelligent Vehicle Heterogeneous Computing","authors":"Liqiang Wang;Meng Wang","doi":"10.1109/OJVT.2025.3634375","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3634375","url":null,"abstract":"The increasing complexity of perception and decision-making tasks in intelligent connected vehicles has driven the evolution of on-board computing platforms toward heterogeneous architectures. However, the dynamic nature of workloads, the need for multi-objective optimization, and stringent safety constraints pose significant challenges to scheduling. To address the limitations of existing approaches in balancing multiple objectives and ensuring safety, this paper proposes a deep reinforcement learning (DRL)-based hierarchical hybrid-action multi-objective adaptive scheduling framework. The framework optimizes latency, energy consumption, reliability, and thermal management by introducing a dynamic weight adjustment mechanism driven by the battery state of charge (SOC) and thermal accumulation. It integrates high-level global task allocation with low-level real-time resource adjustment for adaptive multi-objective trade-offs, while embedding a functional safety fallback mechanism to guarantee hard real-time performance and thermal safety for high-criticality tasks. Experimental results under highway cruising, urban congestion, and high-temperature scenarios show that the proposed method outperforms HEFT, E-List, and Vanilla-DRL in p95 latency, energy consumption, peak temperature, and high-criticality task satisfaction: p95 latency is reduced by 6%–14%, energy consumption by 5%–20%, peak temperature by 2–8°C, and satisfaction rates exceed 97.5%. After model compression, the strategy network achieves inference latency under 5 ms and nearly 40% power reduction on an automotive-grade heterogeneous platform, validating the engineering feasibility of the approach. This work provides a scalable and safety-aware solution for efficient heterogeneous computing scheduling in intelligent vehicles.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"108-123"},"PeriodicalIF":4.8,"publicationDate":"2025-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11251222","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145729518","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-11DOI: 10.1109/OJVT.2025.3631629
Yuhang Chen;Heyin Shen;Chong Han
The evolution of wireless communication toward next-generation networks introduces unprecedented demands on data rates, latency, and connectivity. To meet these requirements, two key trends have emerged: the use of higher communication frequencies to provide broader bandwidth, and the deployment of massive multiple-input multiple-output systems with large antenna arrays to compensate for propagation losses and enhance spatial multiplexing. These advancements significantly extend the Rayleigh distance, enabling near-field (NF) propagation alongside the traditional far-field (FF) regime. As user communication distances dynamically span both FF and NF regions, cross-field (CF) communication has also emerged as a practical consideration. Beam management (BM)—including beam scanning, channel state information estimation, beamforming, and beam tracking—plays a central role in maintaining reliable directional communications. While most existing BM techniques are developed for FF channels, recent works begin to address the unique characteristics of NF and CF regimes. This survey presents a comprehensive review of BM techniques from the perspective of propagation fields. We begin by building the basic through analyzing the modeling of FF, NF, and CF channels, along with the associated beam patterns for alignment. Then, we categorize BM techniques by methodologies, and discuss their operational differences across propagation regimes, highlighting how field-dependent channel characteristics influence design tradeoffs and implementation complexity. In addition, for each BM method, we identify open challenges and future research directions, including extending FF methods to NF/CF scenarios, developing unified BM strategies for field-agnostic deployment, and designing low-overhead BM solutions for dynamic environments.
{"title":"Cross Far- and Near-Field Beam Management Technologies in Millimeter-Wave and Terahertz MIMO Systems","authors":"Yuhang Chen;Heyin Shen;Chong Han","doi":"10.1109/OJVT.2025.3631629","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3631629","url":null,"abstract":"The evolution of wireless communication toward next-generation networks introduces unprecedented demands on data rates, latency, and connectivity. To meet these requirements, two key trends have emerged: the use of higher communication frequencies to provide broader bandwidth, and the deployment of massive multiple-input multiple-output systems with large antenna arrays to compensate for propagation losses and enhance spatial multiplexing. These advancements significantly extend the Rayleigh distance, enabling near-field (NF) propagation alongside the traditional far-field (FF) regime. As user communication distances dynamically span both FF and NF regions, cross-field (CF) communication has also emerged as a practical consideration. Beam management (BM)—including beam scanning, channel state information estimation, beamforming, and beam tracking—plays a central role in maintaining reliable directional communications. While most existing BM techniques are developed for FF channels, recent works begin to address the unique characteristics of NF and CF regimes. This survey presents a comprehensive review of BM techniques from the perspective of propagation fields. We begin by building the basic through analyzing the modeling of FF, NF, and CF channels, along with the associated beam patterns for alignment. Then, we categorize BM techniques by methodologies, and discuss their operational differences across propagation regimes, highlighting how field-dependent channel characteristics influence design tradeoffs and implementation complexity. In addition, for each BM method, we identify open challenges and future research directions, including extending FF methods to NF/CF scenarios, developing unified BM strategies for field-agnostic deployment, and designing low-overhead BM solutions for dynamic environments.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"73-107"},"PeriodicalIF":4.8,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11239413","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145729519","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The Vehicle-to-Ground (V2G) emergency communication network is a dedicated network established to respond to emergencies, such as natural disasters and traffic accidents, and it plays a crucial role in ensuring the safe and smooth operation of vehicles. Composed of numerous devices, this network is inevitably exposed to failure risks due to prolonged operation, complex designs, and insufficient management and maintenance. Faults in network nodes may undermine the reliability of vehicle-to-ground communication. Rapid fault localization is critical to the maintenance and management of network device. However, current localization methods face issues like excessively long probing paths, high localization costs, and low accuracy—all of which lead to subpar performance in real-world fault localization scenarios. To address these problems, we introduce a novel Multi-stage Group Probe (MGP) localization method, designed to balance localization cost and accuracy effectively. Specifically, we first present a network localization model and the concept of "uncertain information volume of network node states," which quantifies the cost and efficiency of localization. Second, leveraging graph theory, we propose the idea of network probing subgraphs and constrain the number of probing stations and probe lengths, while developing algorithms for selecting probing stations and planning probing paths. Additionally, we introduce a group probe localization method that incorporates information feedback to reduce costs. Finally, we evaluate the MGP against other probe localization approaches across different networks. Experimental results demonstrate that MGP outperforms comparative methods in terms of localization cost, accuracy, and efficiency.
{"title":"MGP: Multi-Stage Grouped Probe Detection for Fault Localization in Vehicle-to-Ground Communication Networks","authors":"Wenxiao Wang;Ping Dong;Yuyang Zhang;Wenxuan Qiao;Xiaoya Zhang;Chengxiao Yu;Hongke Zhang","doi":"10.1109/OJVT.2025.3630603","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3630603","url":null,"abstract":"The Vehicle-to-Ground (V2G) emergency communication network is a dedicated network established to respond to emergencies, such as natural disasters and traffic accidents, and it plays a crucial role in ensuring the safe and smooth operation of vehicles. Composed of numerous devices, this network is inevitably exposed to failure risks due to prolonged operation, complex designs, and insufficient management and maintenance. Faults in network nodes may undermine the reliability of vehicle-to-ground communication. Rapid fault localization is critical to the maintenance and management of network device. However, current localization methods face issues like excessively long probing paths, high localization costs, and low accuracy—all of which lead to subpar performance in real-world fault localization scenarios. To address these problems, we introduce a novel Multi-stage Group Probe (MGP) localization method, designed to balance localization cost and accuracy effectively. Specifically, we first present a network localization model and the concept of \"uncertain information volume of network node states,\" which quantifies the cost and efficiency of localization. Second, leveraging graph theory, we propose the idea of network probing subgraphs and constrain the number of probing stations and probe lengths, while developing algorithms for selecting probing stations and planning probing paths. Additionally, we introduce a group probe localization method that incorporates information feedback to reduce costs. Finally, we evaluate the MGP against other probe localization approaches across different networks. Experimental results demonstrate that MGP outperforms comparative methods in terms of localization cost, accuracy, and efficiency.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"27-40"},"PeriodicalIF":4.8,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11235598","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145674876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：