In industrial environments, the wireless infrastructure is functional for offering services such as communication and positioning of industrial assets. However, the frequently occurring Non-Line-of-Sight (NLoS) conditions in industrial scenarios cause the wireless receiver to have positional information from a limited and varying number of wireless transmitters between consecutive time steps, leading to ambiguities in wireless infrastructure-based positioning. In this paper, we propose PosGNN, a novel data fusion solution based on the Graph Neural Network (GNN) approach that allows us to estimate the position of the User Equipment (UE) by fusing the positional information from the available wireless transmitters at each time step with the UE sensor technology. The performance of the proposed method is assessed using an experimental setup of Ultra-Wideband (UWB) technology as wireless infrastructure at $3.7 - text{4.2},text{GHz}$ frequency band, the Inertial Measurement Unit (IMU) as UE-side sensor, and the Automated Guided Vehicle (AGV) as the target UE to be positioned. The experimental results demonstrate the exceptional performance of our approach over the conventional model-based approach, Extended Kalman Filter (EKF), and the data-driven approach, Deep Neural Network (DNN), achieving an average positioning error of less than $text{15},text{cm}$ in harsh industrial environments.
{"title":"PosGNN: A Graph Neural Network Based Multimodal Data Fusion for Indoor Positioning in Industrial Non-Line-of-Sight Scenarios","authors":"Karthik Muthineni;Alexander Artemenko;Daniel Abode;Josep Vidal;Montse Nájar","doi":"10.1109/OJVT.2025.3630970","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3630970","url":null,"abstract":"In industrial environments, the wireless infrastructure is functional for offering services such as communication and positioning of industrial assets. However, the frequently occurring Non-Line-of-Sight (NLoS) conditions in industrial scenarios cause the wireless receiver to have positional information from a limited and varying number of wireless transmitters between consecutive time steps, leading to ambiguities in wireless infrastructure-based positioning. In this paper, we propose PosGNN, a novel data fusion solution based on the Graph Neural Network (GNN) approach that allows us to estimate the position of the User Equipment (UE) by fusing the positional information from the available wireless transmitters at each time step with the UE sensor technology. The performance of the proposed method is assessed using an experimental setup of Ultra-Wideband (UWB) technology as wireless infrastructure at <inline-formula><tex-math>$3.7 - text{4.2},text{GHz}$</tex-math></inline-formula> frequency band, the Inertial Measurement Unit (IMU) as UE-side sensor, and the Automated Guided Vehicle (AGV) as the target UE to be positioned. The experimental results demonstrate the exceptional performance of our approach over the conventional model-based approach, Extended Kalman Filter (EKF), and the data-driven approach, Deep Neural Network (DNN), achieving an average positioning error of less than <inline-formula><tex-math>$text{15},text{cm}$</tex-math></inline-formula> in harsh industrial environments.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"15-26"},"PeriodicalIF":4.8,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11235985","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145612200","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}
Vision-Language Models (VLMs) are becoming increasingly powerful, demonstrating strong performance on tasks that require both visual and textual understanding. Their strong generalisation abilities make them a promising component for automated driving systems, which must handle unexpected corner cases. However, to be trusted in such safety-critical applications, a model must first possess a reliable perception system. Since critical objects and agents in traffic scenes are often at a distance, we require systems that are not “shortsighted,” i.e., systems with strong perception capabilities at both close (up to 20 meters) and long (30+ meters) range. With this in mind, we introduce Distance-Annotated Traffic Perception Question Answering (DTPQA), the first Visual Question Answering (VQA) benchmark focused solely on perception-based questions in traffic scenes, enriched with distance annotations. By excluding questions that require reasoning, we ensure that model performance reflects perception capabilities alone. Since automated driving hardware has limited processing power and cannot support large VLMs, our study centers on smaller VLMs. We evaluate several state-of-the-art (SOTA) small VLMs on DTPQA and show that, despite the simplicity of the questions, these models significantly underperform compared to humans (∼60% average accuracy for the best-performing small VLM versus ∼85% human performance). However, the human sample size was relatively small, which imposes statistical limitations. We also identify specific perception tasks, such as distinguishing left from right, that remain particularly challenging. We hope our findings will encourage further research into improving the perception capabilities of small VLMs in traffic scenarios, making them more suitable for automated driving applications.
{"title":"Evaluating Small Vision-Language Models on Distance-Dependent Traffic Perception","authors":"Nikos Theodoridis;Tim Brophy;Reenu Mohandas;Ganesh Sistu;Fiachra Collins;Anthony Scanlan;Ciarán Eising","doi":"10.1109/OJVT.2025.3629318","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3629318","url":null,"abstract":"Vision-Language Models (VLMs) are becoming increasingly powerful, demonstrating strong performance on tasks that require both visual and textual understanding. Their strong generalisation abilities make them a promising component for automated driving systems, which must handle unexpected corner cases. However, to be trusted in such safety-critical applications, a model must first possess a reliable perception system. Since critical objects and agents in traffic scenes are often at a distance, we require systems that are not “shortsighted,” i.e., systems with strong perception capabilities at both close (up to 20 meters) and long (30+ meters) range. With this in mind, we introduce Distance-Annotated Traffic Perception Question Answering (DTPQA), the first Visual Question Answering (VQA) benchmark focused solely on perception-based questions in traffic scenes, enriched with distance annotations. By excluding questions that require reasoning, we ensure that model performance reflects perception capabilities alone. Since automated driving hardware has limited processing power and cannot support large VLMs, our study centers on smaller VLMs. We evaluate several state-of-the-art (SOTA) small VLMs on DTPQA and show that, despite the simplicity of the questions, these models significantly underperform compared to humans (∼60% average accuracy for the best-performing small VLM versus ∼85% human performance). However, the human sample size was relatively small, which imposes statistical limitations. We also identify specific perception tasks, such as distinguishing left from right, that remain particularly challenging. We hope our findings will encourage further research into improving the perception capabilities of small VLMs in traffic scenarios, making them more suitable for automated driving applications.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"54-72"},"PeriodicalIF":4.8,"publicationDate":"2025-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11230063","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145729342","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 : 2025-11-04DOI: 10.1109/OJVT.2025.3628652
Louis-Romain Joly;Vivien Lacorre;Krister Wolff
Efficient representation and querying of railway networks are crucial for autonomous railway systems and digital infrastructure management. This paper introduces VEctorial Railway NEtwork (VERNE), an interpretable data structure and algorithm that integrates vector-based spatial partitioning with a railway-specific topological framework to enhance network representation and navigation. VERNE is designed to optimize query efficiency, reduce memory footprint, and ensure scalability for real-time applications. Its internal mechanism results from a comparative performance analysis between a $k$-d tree, an STRtree and two custom algorithms, highlighting trade-offs in computational efficiency and memory overhead. The proposed approach is validated using datasets from both the French and Swedish railway networks, demonstrating its effectiveness in real-world scenarios. The results indicate that VERNE provides a robust and scalable solution for railway infrastructure modeling, offering improvements in localization speed and computational efficiency. Another advantage is that it inherently manipulates atomic elements which can contain any information relevant to directly perform navigation onboard an autonomous robot. This work contributes to the advancement of railway digitalization by providing a structured methodology for spatial data processing in autonomous railway systems.
{"title":"VERNE: A Spatial Data Structure Representing Railway Networks for Autonomous Robot Navigation","authors":"Louis-Romain Joly;Vivien Lacorre;Krister Wolff","doi":"10.1109/OJVT.2025.3628652","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3628652","url":null,"abstract":"Efficient representation and querying of railway networks are crucial for autonomous railway systems and digital infrastructure management. This paper introduces <bold>VEctorial Railway NEtwork (VERNE)</b>, an interpretable data structure and algorithm that integrates vector-based spatial partitioning with a railway-specific topological framework to enhance network representation and navigation. VERNE is designed to optimize query efficiency, reduce memory footprint, and ensure scalability for real-time applications. Its internal mechanism results from a comparative performance analysis between a <inline-formula><tex-math>$k$</tex-math></inline-formula>-d tree, an STRtree and two custom algorithms, highlighting trade-offs in computational efficiency and memory overhead. The proposed approach is validated using datasets from both the French and Swedish railway networks, demonstrating its effectiveness in real-world scenarios. The results indicate that VERNE provides a robust and scalable solution for railway infrastructure modeling, offering improvements in localization speed and computational efficiency. Another advantage is that it inherently manipulates atomic elements which can contain any information relevant to directly perform navigation onboard an autonomous robot. This work contributes to the advancement of railway digitalization by providing a structured methodology for spatial data processing in autonomous railway systems.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"1-14"},"PeriodicalIF":4.8,"publicationDate":"2025-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11224796","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145595123","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 : 2025-10-30DOI: 10.1109/OJVT.2025.3627139
Priyanka Maity;Deepika Harish;Suraj Srivastava;Aditya K. Jagannatham;Lajos Hanzo
With the growing demand for integrated sensing and communication (ISAC) in next-generation wireless networks, efficient target localization techniques conceived for mmWave MIMO systems have becomeincreasingly important. In this context, we propose a Sparse Bayesian Learning (SBL)-aided extended target localization framework for orthogonal frequency division multiplexing (OFDM)-based mmWave MIMO systems. The proposed approach explicitly considers the impact of intercarrier interference (ICI) arising in mobile environments, which is often overlooked in conventional schemes. Our framework is designed for hybrid mmWave MIMO architectures, where the number of radio frequency (RF) chains is considerably lower than the number of antennas, ensuring hardware efficiency. To achieve high-precision target localization, we introduce a delay, Doppler, and angular (DDA)-domain representation of the target, enabling accurate estimation of target parameters. The proposed algorithm leverages the inherent three-dimensional (3D) sparsity in the DDA domain of the scattering environment and employs the powerful SBL framework for effective parameter estimation. Furthermore, to address practical scenarios where the actual target parameters may not align with finite-resolution grids, we develop an enhanced off-grid SBL (OSBL) method based on super-resolution principles. This recursive grid refinement approach progressively improves the estimation accuracy. Additionally, we derive the Cramér-Rao bound (CRB) and Bayesian CRB to theoretically characterize the achievable estimation performance. Simulation results confirm that the proposed method significantly outperforms existing algorithms in terms of estimation accuracy and robustness.
{"title":"Super-Resolution-Based Bayesian Learning for the Localization of Extended Targets in mmWave MIMO OFDM Systems","authors":"Priyanka Maity;Deepika Harish;Suraj Srivastava;Aditya K. Jagannatham;Lajos Hanzo","doi":"10.1109/OJVT.2025.3627139","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3627139","url":null,"abstract":"With the growing demand for integrated sensing and communication (ISAC) in next-generation wireless networks, efficient target localization techniques conceived for mmWave MIMO systems have becomeincreasingly important. In this context, we propose a Sparse Bayesian Learning (SBL)-aided extended target localization framework for orthogonal frequency division multiplexing (OFDM)-based mmWave MIMO systems. The proposed approach explicitly considers the impact of intercarrier interference (ICI) arising in mobile environments, which is often overlooked in conventional schemes. Our framework is designed for hybrid mmWave MIMO architectures, where the number of radio frequency (RF) chains is considerably lower than the number of antennas, ensuring hardware efficiency. To achieve high-precision target localization, we introduce a delay, Doppler, and angular (DDA)-domain representation of the target, enabling accurate estimation of target parameters. The proposed algorithm leverages the inherent three-dimensional (3D) sparsity in the DDA domain of the scattering environment and employs the powerful SBL framework for effective parameter estimation. Furthermore, to address practical scenarios where the actual target parameters may not align with finite-resolution grids, we develop an enhanced off-grid SBL (OSBL) method based on super-resolution principles. This recursive grid refinement approach progressively improves the estimation accuracy. Additionally, we derive the Cramér-Rao bound (CRB) and Bayesian CRB to theoretically characterize the achievable estimation performance. Simulation results confirm that the proposed method significantly outperforms existing algorithms in terms of estimation accuracy and robustness.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"6 ","pages":"3000-3016"},"PeriodicalIF":4.8,"publicationDate":"2025-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11222915","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560673","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 : 2025-10-22DOI: 10.1109/OJVT.2025.3623883
Hugo Hawkins;Chao Xu;Lie-Liang Yang;Lajos Hanzo
Affine Frequency Division Multiplexing (AFDM) has attracted substantial research interest due to its implementational similarity to Orthogonal Frequency-Division Multiplexing (OFDM), whilst attaining comparable performance to Orthogonal Time Frequency Space (OTFS). Hence, we embark on an in-depth performance characterisation of coded AFDM and of its equivalent OTFS counterpart. Soft-Minimum Mean Square Error (MMSE) reception taking into account a priori probabilities in the weighting matrix is applied in conjunction with Recursive Systematic Convolutional (RSC)- and RSCUnity Rate Convolutional (URC) coding to AFDM. Iterative decoding convergence analysis is carried out with the aid of the powerful semi-analytical tool of EXtrinsic Information Transfer (EXIT) chart, and its Bit Error Rate (BER) performance is compared to OFDM and to the equivalent OTFS configurations. As there are no consistent configurations of AFDM and OTFS utilised in the literature to compare their relative performances, two AFDM configurations and three OTFS configurations are considered. The results show that the BER of AFDM is lower than that of the equivalent OTFS configurations at high Energy per bit over Noise power (E$_{mathrm{{b}}}$/N$_{0}$) for small system matrix dimensions, for a low number of iterations, and for high code rates. In all other cases, the BER of AFDM is shown to be similar to that of its equivalent OTFS configurations. Given that the RSC BER performance fails to improve beyond two iterations, this solution is recommended for low-complexity transceivers. By contrast, if the extra complexity of the RSC-URC aided transceiver is affordable, an extra (E$_{mathrm{{b}}}$/N$_{0}$) gain of 1.8 dB may be attained at a BER of $10^{-5}$ and a code rate of 0.5.
仿射频分复用（AFDM）由于其实现与正交频分复用（OFDM）相似，同时获得与正交时频空间（OTFS）相当的性能而引起了大量的研究兴趣。因此，我们着手对编码AFDM及其等效OTFS对立物进行深入的性能表征。考虑到加权矩阵中的先验概率的软最小均方误差（MMSE）接收与递归系统卷积（RSC）和RSCUnity Rate卷积（URC）编码一起应用于AFDM。借助强大的外部信息传输（EXtrinsic Information Transfer， EXIT）图半分析工具进行了迭代译码收敛分析，并将其误码率（BER）性能与OFDM和等效OTFS配置进行了比较。由于在文献中没有使用一致的AFDM和OTFS配置来比较它们的相对性能，因此考虑两种AFDM配置和三种OTFS配置。结果表明，当系统矩阵维数小、迭代次数少、码率高时，AFDM的误码率比等效OTFS配置的误码率低。在所有其他情况下，AFDM的误码率显示与其等效OTFS配置的误码率相似。考虑到RSC误码率的性能在两次迭代之后就无法提高，建议将此解决方案用于低复杂度的收发器。相比之下，如果RSC-URC辅助收发器的额外复杂性是可以承受的，则可以在BER为10^{-5}$和码率为0.5的情况下获得1.8 dB的额外增益（E$_{ mathm {{b}}}$/N$_{0}$）。
{"title":"Iterative Soft-MMSE Detection Aided AFDM and OTFS","authors":"Hugo Hawkins;Chao Xu;Lie-Liang Yang;Lajos Hanzo","doi":"10.1109/OJVT.2025.3623883","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3623883","url":null,"abstract":"Affine Frequency Division Multiplexing (AFDM) has attracted substantial research interest due to its implementational similarity to Orthogonal Frequency-Division Multiplexing (OFDM), whilst attaining comparable performance to Orthogonal Time Frequency Space (OTFS). Hence, we embark on an in-depth performance characterisation of coded AFDM and of its equivalent OTFS counterpart. Soft-Minimum Mean Square Error (MMSE) reception taking into account <italic>a priori</i> probabilities in the weighting matrix is applied in conjunction with Recursive Systematic Convolutional (RSC)- and RSCUnity Rate Convolutional (URC) coding to AFDM. Iterative decoding convergence analysis is carried out with the aid of the powerful semi-analytical tool of EXtrinsic Information Transfer (EXIT) chart, and its Bit Error Rate (BER) performance is compared to OFDM and to the equivalent OTFS configurations. As there are no consistent configurations of AFDM and OTFS utilised in the literature to compare their relative performances, two AFDM configurations and three OTFS configurations are considered. The results show that the BER of AFDM is lower than that of the equivalent OTFS configurations at high Energy per bit over Noise power (E<inline-formula><tex-math>$_{mathrm{{b}}}$</tex-math></inline-formula>/N<inline-formula><tex-math>$_{0}$</tex-math></inline-formula>) for small system matrix dimensions, for a low number of iterations, and for high code rates. In all other cases, the BER of AFDM is shown to be similar to that of its equivalent OTFS configurations. Given that the RSC BER performance fails to improve beyond two iterations, this solution is recommended for low-complexity transceivers. By contrast, if the extra complexity of the RSC-URC aided transceiver is affordable, an extra (E<inline-formula><tex-math>$_{mathrm{{b}}}$</tex-math></inline-formula>/N<inline-formula><tex-math>$_{0}$</tex-math></inline-formula>) gain of 1.8 dB may be attained at a BER of <inline-formula><tex-math>$10^{-5}$</tex-math></inline-formula> and a code rate of 0.5.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"6 ","pages":"2944-2959"},"PeriodicalIF":4.8,"publicationDate":"2025-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11214369","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145510082","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 : 2025-10-20DOI: 10.1109/OJVT.2025.3623913
Adam Weaver;Annette von Jouanne;Douglas Sicker;Alex Yokochi
Modern vehicles are increasingly more cyber-physical as well as more connected with each passing year. Manufacturers are innovating new technologies (including performance, automation, comfort, and safety) that enhance the driver/passenger experience and continue to move towards increased automation. However, each of these cyber-physical systems poses a potential additional vulnerable attack surface for malicious actors to exploit. Due to high costs, safety risks, and logistical difficulties of testing full vehicles in motion, most research in assessing the cybersecurity of vehicles has focused on simulation, vehicle subsystem(s), or constricted case studies, and not real-world vehicle testing and realistic human interaction assessment. To address this shortcoming, hardware-in-the-loop (HIL) cyber vulnerability testing of fully operational vehicles is needed. This paper presents a review of vehicle cybersecurity research and testing including common technical, logistical, and human factors issues as well as current regulations and guidance. Informative research-focused case studies are presented followed by a proposed cybersecurity vehicle-in-the-loop testbed integrated with a dynamometer to provide a comprehensive, robust, and safe test environment where true effects of cyber testing can be evaluated on a complete vehicle.
{"title":"Cybersecurity Dynamometer Testbed: A Review to Advance Vehicle-in-the-Loop Testing of Traditional, Connected and Autonomous Vehicles","authors":"Adam Weaver;Annette von Jouanne;Douglas Sicker;Alex Yokochi","doi":"10.1109/OJVT.2025.3623913","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3623913","url":null,"abstract":"Modern vehicles are increasingly more cyber-physical as well as more connected with each passing year. Manufacturers are innovating new technologies (including performance, automation, comfort, and safety) that enhance the driver/passenger experience and continue to move towards increased automation. However, each of these cyber-physical systems poses a potential additional vulnerable attack surface for malicious actors to exploit. Due to high costs, safety risks, and logistical difficulties of testing full vehicles in motion, most research in assessing the cybersecurity of vehicles has focused on simulation, vehicle subsystem(s), or constricted case studies, and not real-world vehicle testing and realistic human interaction assessment. To address this shortcoming, hardware-in-the-loop (HIL) cyber vulnerability testing of fully operational vehicles is needed. This paper presents a review of vehicle cybersecurity research and testing including common technical, logistical, and human factors issues as well as current regulations and guidance. Informative research-focused case studies are presented followed by a proposed cybersecurity vehicle-in-the-loop testbed integrated with a dynamometer to provide a comprehensive, robust, and safe test environment where true effects of cyber testing can be evaluated on a complete vehicle.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"6 ","pages":"2925-2943"},"PeriodicalIF":4.8,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11208586","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145455949","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}
Hardware-in-the-Loop (HiL) driving simulators are valuable tools in assuring efficiency throughout the entire vehicle development and life cycle. Nonetheless, these techniques are accused to prejudice the simulator flexibility with respect to more conventional Model-in-the-Loop techniques. The vehicle network integration activity required to set off the HiL simulator, known as restbus simulation, is perceived as the main source of these perplexities. All vehicle signals to be emulated, defining the interface between real and simulated world, represents a pivotal point in simulator contexts and should be addressed methodically from the start, prior to every other decisional process, since they affect simulator architecture possibilities and limitations.The present article proposes a method to evaluate an integration task and synthetize key aspects of the vehicle communication network, supporting both high and low-level decisions in software, model and validation plans development. Thanks to the abstraction of the communication protocol formalism, integration activities can be conformed at a level of detail suitable for typical simulator engineers’ educational backgrounds, reducing working effort, time and expertise required to update any driving simulator to specific HiL setups. The entire method is presented and discussed; electing an application case, the value of the approach is highlighted and the reasoning for its outputs is explained both practically and conceptually. Advantages and limitations of the proposed approaches are hence discussed emphasizing its effects on numerical model development, programming activities and operational workflow rationalisation.
{"title":"Hardware-in-the-Loop Driving Simulators: Simplifying Real Component Integration in Simulated Environments","authors":"Alessio Anticaglia;Renzo Capitani;Claudio Annicchiarico","doi":"10.1109/OJVT.2025.3622519","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3622519","url":null,"abstract":"Hardware-in-the-Loop (HiL) driving simulators are valuable tools in assuring efficiency throughout the entire vehicle development and life cycle. Nonetheless, these techniques are accused to prejudice the simulator flexibility with respect to more conventional Model-in-the-Loop techniques. The vehicle network integration activity required to set off the HiL simulator, known as restbus simulation, is perceived as the main source of these perplexities. All vehicle signals to be emulated, defining the interface between real and simulated world, represents a pivotal point in simulator contexts and should be addressed methodically from the start, prior to every other decisional process, since they affect simulator architecture possibilities and limitations.The present article proposes a method to evaluate an integration task and synthetize key aspects of the vehicle communication network, supporting both high and low-level decisions in software, model and validation plans development. Thanks to the abstraction of the communication protocol formalism, integration activities can be conformed at a level of detail suitable for typical simulator engineers’ educational backgrounds, reducing working effort, time and expertise required to update any driving simulator to specific HiL setups. The entire method is presented and discussed; electing an application case, the value of the approach is highlighted and the reasoning for its outputs is explained both practically and conceptually. Advantages and limitations of the proposed approaches are hence discussed emphasizing its effects on numerical model development, programming activities and operational workflow rationalisation.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"6 ","pages":"2891-2908"},"PeriodicalIF":4.8,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11205838","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145456067","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}
Reliable perception in automated vehicles under adverse conditions, such as fog, rain, snow, and lens defocus, is essential for maintaining the safety of road actors and particularly of vulnerable road users. While prior work has primarily focused on camera occlusions, the impact on RADAR and LiDAR remains underexplored, particularly in a unified Bird's Eye View (BEV) space. To address this gap, we first apply occlusion to all three primary sensors: camera, RADAR, and LiDAR, and then systematically investigate its impact by projecting their outputs into the BEV space for unified analysis of vehicle and map segmentation. A parametrised occlusion pipeline is developed to apply occlusions to each of the sensor modalities. We evaluate both geometry-based and transformer-based fusion architectures, revealing that transformer-based architectures consistently demonstrate greater robustness to sensor degradation. Notably, we demonstrate that BEVCar achieves 45.6% vehicle Intersection-over-Union (IoU) and 53.6% Mean Intersection-over-Union (mIoU) under camera occlusion, surpassing other State-of-the-art (SOTA) models such as MMTraP (37.9% IoU / 47.9% mIoU) and CVT (36.0% IoU / 46.6% mIoU). These improvements are statistically significant (paired t-tests with 95% CI bootstrap, $p < 0.001$). Furthermore, projecting camera features into the BEV space using a backward projection strategy seems to offer greater resilience to occlusion than forward projection. These insights highlight the importance of architectural design, projection choice, and multi-sensor fusion in developing robust perception systems for automated driving under realistic multi-sensor occlusions.
在雾、雨、雪和镜头散焦等不利条件下，自动驾驶汽车的可靠感知对于维护道路行为者，特别是弱势道路使用者的安全至关重要。虽然之前的工作主要集中在相机遮挡上，但对雷达和激光雷达的影响仍未得到充分探讨，特别是在统一的鸟瞰图（BEV）空间。为了解决这一差距，我们首先将遮挡应用于所有三个主要传感器：摄像头、雷达和激光雷达，然后通过将它们的输出投影到BEV空间中，系统地研究其影响，以统一分析车辆和地图分割。一个参数化的遮挡管道被开发应用于每个传感器模态的遮挡。我们评估了基于几何的和基于变压器的融合架构，发现基于变压器的架构对传感器退化具有更强的鲁棒性。值得注意的是，我们证明了BEVCar在相机遮挡下实现了45.6%的车辆过路口（IoU）和53.6%的平均过路口（mIoU），超过了其他最先进的（SOTA）模型，如MMTraP （37.9% IoU / 47.9% mIoU）和CVT （36.0% IoU / 46.6% mIoU）。这些改进在统计上是显著的（配对t检验与95% CI bootstrap， p < 0.001）。此外，使用向后投影策略将相机特征投影到BEV空间中，似乎比向前投影提供了更大的抗遮挡能力。这些见解强调了建筑设计、投影选择和多传感器融合在开发现实多传感器遮挡下自动驾驶的鲁棒感知系统中的重要性。
{"title":"Exploring Sensor Impact and Architectural Robustness in Adverse Weather on BEV Perception","authors":"Sanjay Kumar;Sushil Sharma;Rabia Asghar;Reenu Mohandas;Tim Brophy;Ganesh Sistu;Eoin Martino Grua;Valentina Donzella;Ciarán Eising","doi":"10.1109/OJVT.2025.3621862","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3621862","url":null,"abstract":"Reliable perception in automated vehicles under adverse conditions, such as fog, rain, snow, and lens defocus, is essential for maintaining the safety of road actors and particularly of vulnerable road users. While prior work has primarily focused on camera occlusions, the impact on RADAR and LiDAR remains underexplored, particularly in a unified Bird's Eye View (BEV) space. To address this gap, we first apply occlusion to all three primary sensors: camera, RADAR, and LiDAR, and then systematically investigate its impact by projecting their outputs into the BEV space for unified analysis of vehicle and map segmentation. A parametrised occlusion pipeline is developed to apply occlusions to each of the sensor modalities. We evaluate both geometry-based and transformer-based fusion architectures, revealing that transformer-based architectures consistently demonstrate greater robustness to sensor degradation. Notably, we demonstrate that BEVCar achieves 45.6% vehicle Intersection-over-Union (IoU) and 53.6% Mean Intersection-over-Union (mIoU) under camera occlusion, surpassing other State-of-the-art (SOTA) models such as MMTraP (37.9% IoU / 47.9% mIoU) and CVT (36.0% IoU / 46.6% mIoU). These improvements are statistically significant (paired t-tests with 95% CI bootstrap, <inline-formula><tex-math>$p < 0.001$</tex-math></inline-formula>). Furthermore, projecting camera features into the BEV space using a backward projection strategy seems to offer greater resilience to occlusion than forward projection. These insights highlight the importance of architectural design, projection choice, and multi-sensor fusion in developing robust perception systems for automated driving under realistic multi-sensor occlusions.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"6 ","pages":"2857-2875"},"PeriodicalIF":4.8,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11204511","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145455924","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 : 2025-10-09DOI: 10.1109/OJVT.2025.3619823
Sachin Janardhanan;Jonas Persson;Mats Jonasson;Bengt Jacobson;Esteban R Gelso;Leon Henderson
This paper proposes an energy efficient hierarchical wheel torque controller for a 4 × 4 heavy electric vehicle equipped with multiple electric drivetrains. The controller consists of two main components: a global force reference generator and a control allocator. The global force reference generator computes motion requests based on steering wheel angle and longitudinal acceleration inputs, while adhering to actuator and tire force constraints. For this purpose, a linear time-varying model predictive controller (LTV-MPC) is employed to minimize the squared errors in yaw rate and longitudinal acceleration over a short prediction horizon. Concurrently, the controller dynamically identifies safe operating limits based on current driving conditions. These limits are then used to adjust the state cost weights dynamically, thereby improving the effectiveness of the MPC cost function. The control allocator (CA) subsequently distributes the force demands from the global reference generator among the electric machines and friction brakes. This allocation process minimizes instantaneous power losses while respecting actuator and tire force constraints. To further enhance energy efficiency, the method leverages the heterogeneous nature of the electric machines by minimizing not only operational power losses but also idle losses (power losses at zero torque), ensuring safe vehicle operation. The proposed strategy is evaluated using a high-fidelity vehicle model under various driving scenarios, including low-friction surfaces and near-handling-limit conditions. Simulation results demonstrate that dynamically varying state cost weights in conjunction with safe operating limits significantly improves vehicle performance, enhances energy efficiency, and reduces driver effort.
{"title":"Energy-Efficient Wheel Torque Distribution for Heavy Electric Vehicles With Adaptive Model Predictive Control and Control Allocation","authors":"Sachin Janardhanan;Jonas Persson;Mats Jonasson;Bengt Jacobson;Esteban R Gelso;Leon Henderson","doi":"10.1109/OJVT.2025.3619823","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3619823","url":null,"abstract":"This paper proposes an energy efficient hierarchical wheel torque controller for a 4 × 4 heavy electric vehicle equipped with multiple electric drivetrains. The controller consists of two main components: a global force reference generator and a control allocator. The global force reference generator computes motion requests based on steering wheel angle and longitudinal acceleration inputs, while adhering to actuator and tire force constraints. For this purpose, a linear time-varying model predictive controller (LTV-MPC) is employed to minimize the squared errors in yaw rate and longitudinal acceleration over a short prediction horizon. Concurrently, the controller dynamically identifies safe operating limits based on current driving conditions. These limits are then used to adjust the state cost weights dynamically, thereby improving the effectiveness of the MPC cost function. The control allocator (CA) subsequently distributes the force demands from the global reference generator among the electric machines and friction brakes. This allocation process minimizes instantaneous power losses while respecting actuator and tire force constraints. To further enhance energy efficiency, the method leverages the heterogeneous nature of the electric machines by minimizing not only operational power losses but also idle losses (power losses at zero torque), ensuring safe vehicle operation. The proposed strategy is evaluated using a high-fidelity vehicle model under various driving scenarios, including low-friction surfaces and near-handling-limit conditions. Simulation results demonstrate that dynamically varying state cost weights in conjunction with safe operating limits significantly improves vehicle performance, enhances energy efficiency, and reduces driver effort.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"6 ","pages":"2909-2924"},"PeriodicalIF":4.8,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11197915","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145456066","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 : 2025-10-09DOI: 10.1109/OJVT.2025.3619828
Syed Aizaz ul Haq;Muhammad Farhan;Nadir Shah;Fazal Hameed;Gabriel-Miro Muntean
This paper proposes a Vehicle Trajectory-aware Offloading Multi-Objective Optimization Algorithm (VT-MOOA), a multi-objective optimization algorithm that employs energy consumption, communication and computation delays, vehicle trajectory prediction, task division into sub-tasks, and SDN-based load balancing to optimize task offloading from vehicles to suitable edge servers in vehicular edge networks. The main aim of this work is to design an offloading framework that is robust to high vehicle mobility while ensuring energy efficiency, reduced delays, and balanced resource utilization. The proposed VT-MOOA enhances the S-Metric Selection Evolutionary Multi-Objective Algorithm (SMS-EMOA) by integrating hypervolume-based selection for faster convergence and improves solution quality by minimizing computation delay, minimizing transmission energy, and minimizing physical distance of the vehicle from the RSU while satisfying load balancing constraints, thereby efficiently managing resources in highly dynamic vehicular environments. Existing approaches are often slow, provide sub-optimal solutions due to single objective, false positive prediction or crowding distance reliance, and ignore critical parameters such as real-time vehicle mobility and trajectory prediction. The proposed VT-MOOA approach addresses these gaps by considering these important parameters along with energy efficiency, task deadlines, and load balancing, enabling more effective offloading decisions. Extensive simulations with real-world vehicular mobility datasets demonstrate that VT-MOOA achieves 14% lower energy consumption, 11% faster task completion time, and 13% reduction in computation delay, while also improving load distribution by about 17% compared to existing solutions, outperforming them.
{"title":"VT-MOOA: A Vehicle Trajectory-Aware Multi-Objective Optimization Algorithm for Task Offloading in SDN-Based Vehicular Edge Networks","authors":"Syed Aizaz ul Haq;Muhammad Farhan;Nadir Shah;Fazal Hameed;Gabriel-Miro Muntean","doi":"10.1109/OJVT.2025.3619828","DOIUrl":"https://doi.org/10.1109/OJVT.2025.3619828","url":null,"abstract":"This paper proposes a Vehicle Trajectory-aware Offloading Multi-Objective Optimization Algorithm (VT-MOOA), a multi-objective optimization algorithm that employs energy consumption, communication and computation delays, vehicle trajectory prediction, task division into sub-tasks, and SDN-based load balancing to optimize task offloading from vehicles to suitable edge servers in vehicular edge networks. The main aim of this work is to design an offloading framework that is robust to high vehicle mobility while ensuring energy efficiency, reduced delays, and balanced resource utilization. The proposed VT-MOOA enhances the S-Metric Selection Evolutionary Multi-Objective Algorithm (SMS-EMOA) by integrating hypervolume-based selection for faster convergence and improves solution quality by minimizing computation delay, minimizing transmission energy, and minimizing physical distance of the vehicle from the RSU while satisfying load balancing constraints, thereby efficiently managing resources in highly dynamic vehicular environments. Existing approaches are often slow, provide sub-optimal solutions due to single objective, false positive prediction or crowding distance reliance, and ignore critical parameters such as real-time vehicle mobility and trajectory prediction. The proposed <bold>VT-MOOA</b> approach addresses these gaps by considering these important parameters along with energy efficiency, task deadlines, and load balancing, enabling more effective offloading decisions. Extensive simulations with real-world vehicular mobility datasets demonstrate that <bold>VT-MOOA</b> achieves 14% lower energy consumption, 11% faster task completion time, and 13% reduction in computation delay, while also improving load distribution by about 17% compared to existing solutions, outperforming them.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"6 ","pages":"2960-2987"},"PeriodicalIF":4.8,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11197650","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145510081","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：