Pub Date : 2026-01-28DOI: 10.1109/OJVT.2026.3658986
Oumaima Skalli;Sergio Rodriguez Florez;Abdelhafid El Ouardi;Stefano Masi
Due to technological advances in the automotive field, advanced driver assistance systems have attracted increasing interest from various research and development entities. Predicting road users' future trajectories remains an active research challenge for advanced driver assistance systems. Accurate Trajectory Prediction (TP) allows anticipation of surrounding road users' future motion, enabling timely safety-critical interventions such as speed regulation and emergency braking in unexpected driving situations. Recent advances in TP methods based on artificial intelligence have demonstrated remarkably accurate results compared to traditional methods. However, many of these models require a high computational burden, which makes their deployment on embedded architectures with constrained resources challenging. To overcome these constraints, TP models need to be lightweight and efficient to meet the real-time and power consumption requirements of advanced driver assistance systems. In other words, they must maintain high accuracy while guaranteeing low computational load and rapid inference. This paper presents a comparative and experimental review of state-of-the-art vehicle TP models. First, we propose a new taxonomy based on the operating environment, the trajectory output type, and the employed modeling approach to classify existing methods. Then, we evaluate representative approaches w.r.t the taxonomy in terms of accuracy, model complexity, computational performance, and real-time feasibility across a high-performance architecture and an embedded architecture. Finally, we discuss the evaluation results and present key conclusions and future directions.
{"title":"Single-Vehicle Trajectory Prediction: A Review and Experimental Embedded Assessment","authors":"Oumaima Skalli;Sergio Rodriguez Florez;Abdelhafid El Ouardi;Stefano Masi","doi":"10.1109/OJVT.2026.3658986","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3658986","url":null,"abstract":"Due to technological advances in the automotive field, advanced driver assistance systems have attracted increasing interest from various research and development entities. Predicting road users' future trajectories remains an active research challenge for advanced driver assistance systems. Accurate Trajectory Prediction (TP) allows anticipation of surrounding road users' future motion, enabling timely safety-critical interventions such as speed regulation and emergency braking in unexpected driving situations. Recent advances in TP methods based on artificial intelligence have demonstrated remarkably accurate results compared to traditional methods. However, many of these models require a high computational burden, which makes their deployment on embedded architectures with constrained resources challenging. To overcome these constraints, TP models need to be lightweight and efficient to meet the real-time and power consumption requirements of advanced driver assistance systems. In other words, they must maintain high accuracy while guaranteeing low computational load and rapid inference. This paper presents a comparative and experimental review of state-of-the-art vehicle TP models. First, we propose a new taxonomy based on the operating environment, the trajectory output type, and the employed modeling approach to classify existing methods. Then, we evaluate representative approaches w.r.t the taxonomy in terms of accuracy, model complexity, computational performance, and real-time feasibility across a high-performance architecture and an embedded architecture. Finally, we discuss the evaluation results and present key conclusions and future directions.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"639-658"},"PeriodicalIF":4.8,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11366921","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146175620","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}
In this paper, we propose and develop a fifth-generation mobile communication system (5G) emulator that can evaluate the entire downlink (DL) communication of a 5G between multiple base stations (BSs) and multiple user equipments (UEs), including the effect of multicell interference in a virtual cyberspace without conducting outdoor experiments. For implementation, we used the 5G development platform based on OpenAirInterface (5G-OAI) running on a Linux machine. However, the default 5G-OAI cannot construct a system capable of evaluating the effects of multicell interference. To evaluate the impact of multicell interference using a 5G-OAI-based emulator, if a straightforward implementation is to be used, it is necessary to expand the emulator to run many BSs in parallel. However, achieving this on a Linux machine requires a significant amount of additional computation, making it impossible to implement. In the proposed 5G emulator, we implemented a mechanism for generating pseudo-interference signals, thereby achieving the effects of multicell interference with ultra-low computational cost (reduction of more than 98%). We also evaluated the block error rate (BLER) characteristics in the 3GPP urban macro scenario, a multicell environment specified by 3GPP, and demonstrated that we can emulate BLER with a root-mean-square error of approximately 0.03.
{"title":"5G Wireless Emulator for Evaluating Downlink Communication Under Multicell Interference","authors":"Takehito Narukawa;Kazuki Takeda;Keiichi Mizutani;Hiroshi Harada","doi":"10.1109/OJVT.2026.3656603","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3656603","url":null,"abstract":"In this paper, we propose and develop a fifth-generation mobile communication system (5G) emulator that can evaluate the entire downlink (DL) communication of a 5G between multiple base stations (BSs) and multiple user equipments (UEs), including the effect of multicell interference in a virtual cyberspace without conducting outdoor experiments. For implementation, we used the 5G development platform based on OpenAirInterface (5G-OAI) running on a Linux machine. However, the default 5G-OAI cannot construct a system capable of evaluating the effects of multicell interference. To evaluate the impact of multicell interference using a 5G-OAI-based emulator, if a straightforward implementation is to be used, it is necessary to expand the emulator to run many BSs in parallel. However, achieving this on a Linux machine requires a significant amount of additional computation, making it impossible to implement. In the proposed 5G emulator, we implemented a mechanism for generating pseudo-interference signals, thereby achieving the effects of multicell interference with ultra-low computational cost (reduction of more than 98%). We also evaluated the block error rate (BLER) characteristics in the 3GPP urban macro scenario, a multicell environment specified by 3GPP, and demonstrated that we can emulate BLER with a root-mean-square error of approximately 0.03.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"766-780"},"PeriodicalIF":4.8,"publicationDate":"2026-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11360292","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299521","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 : 2026-01-21DOI: 10.1109/OJVT.2026.3656502
Amit Singh;Sanjeev Sharma;Mohit Kumar Sharma;Kuntal Deka;Daniel Benevides da Costa
Orthogonal time frequency space (OTFS) modulation is a promising approach to improve the performance of millimeter wave (mmWave) communication systems at high mobility while leveraging the wide available bandwidth. However, the high mobility and frequencies in the mmWave regime increase the sensitivity of transceivers to hardware impairments (HIs) such as in-phase and quadrature (IQ) imbalance and direct current (DC) offset, degrading the OTFS performance. We develop an unsupervised deep learning (DL)-based approach to learn a hybrid precoder for a mmWave multi-user (MU) multiple-input and multiple-output (MIMO)-OTFS system, referred to as hybrid beamforming MIMO OTFS (HM-OTFS). In addition, a convolutional neural network (CNN)-based signal detector is proposed for the HM-OTFS system to mitigate the impact of HIs. Our results show that the proposed DL-based beamforming (DLBF) outperforms conventional hybrid beamforming (HBF) schemes aided with estimation and compensation of HIs, providing a performance improvement of more than 2 dB. Furthermore, the proposed CNN-based detector provides a huge performance improvement, compared to conventional minimum mean square error (MMSE) and message passing algorithm (MPA) based detectors, even in the presence of imperfect channel state information (CSI). Extensive simulations establish the bit error rate (BER) performance of the proposed schemes in the presence of HIs, with variations in parameters such as number of users, user's mobility, HIs characteristics, and MIMO configuration.
{"title":"Enhanced Hybrid Beamforming and Signal Detection of OTFS in Presence of Hardware Impairments","authors":"Amit Singh;Sanjeev Sharma;Mohit Kumar Sharma;Kuntal Deka;Daniel Benevides da Costa","doi":"10.1109/OJVT.2026.3656502","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3656502","url":null,"abstract":"Orthogonal time frequency space (OTFS) modulation is a promising approach to improve the performance of millimeter wave (mmWave) communication systems at high mobility while leveraging the wide available bandwidth. However, the high mobility and frequencies in the mmWave regime increase the sensitivity of transceivers to hardware impairments (HIs) such as in-phase and quadrature (IQ) imbalance and direct current (DC) offset, degrading the OTFS performance. We develop an <italic>unsupervised</i> deep learning (DL)-based approach to learn a hybrid precoder for a mmWave multi-user (MU) multiple-input and multiple-output (MIMO)-OTFS system, referred to as hybrid beamforming MIMO OTFS (HM-OTFS). In addition, a convolutional neural network (CNN)-based signal detector is proposed for the HM-OTFS system to mitigate the impact of HIs. Our results show that the proposed DL-based beamforming (DLBF) outperforms conventional hybrid beamforming (HBF) schemes aided with estimation and compensation of HIs, providing a performance improvement of more than 2 dB. Furthermore, the proposed CNN-based detector provides a huge performance improvement, compared to conventional minimum mean square error (MMSE) and message passing algorithm (MPA) based detectors, even in the presence of imperfect channel state information (CSI). Extensive simulations establish the bit error rate (BER) performance of the proposed schemes in the presence of HIs, with variations in parameters such as number of users, user's mobility, HIs characteristics, and MIMO configuration.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"582-597"},"PeriodicalIF":4.8,"publicationDate":"2026-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11359726","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146175608","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 : 2026-01-21DOI: 10.1109/OJVT.2026.3656394
Inam Ullah;Hesham El-Sayed;Alexis Dowhuszko;Manzoor Ahmed Khan;Jyri Hämäläinen
The rise of data-intensive applications in Fifth-Generation (5G) mobile networks demands that next-generation mobile systems deliver seamless, high-bandwidth, and immersive services with improved quality of service. To address these challenges, the use of Integrated Access and Backhaul (IAB) nodes operating over millimeter-wave frequency bands onboard Unmanned Aerial Vehicle (UAV) presents a promising solution. The UAV-mounted IAB network has the potential to enhance line-of-sight conditions to the donor Base Station (BS) via the backhaul link, enabling temporary high data rates in mission-critical and emergency response communication scenarios that require a rapid deployment of new network elements for boosting cellular coverage. This article proposed a framework that integrates terrestrial IAB nodes, non-terrestrial UAV-mounted IAB nodes, and terrestrial donor BS, operating in a dense urban Manhattan-like environment. The research work primarily focuses on how variations in UAV-mounted IAB altitudes, donor BS down-tilt angle, and IAB antenna configuration influence the downlink end-to-end (E2E) spectral efficiency performance of mobile users. Simulation results demonstrate that significant performance gains can be achieved when non-terrestrial IAB nodes are deployed at suitable altitudes when equipped with appropriate antenna configurations. These improvements are further improved when the donor BS employs properly adjusted down-tilt angles, enabling the hybrid terrestrial-aerial IAB mobile network to operate more efficiently and deliver enhanced E2E performances.
{"title":"Performance Evaluation of Non-Terrestrial IAB Nodes at Varying Altitudes in Dense Urban Environments","authors":"Inam Ullah;Hesham El-Sayed;Alexis Dowhuszko;Manzoor Ahmed Khan;Jyri Hämäläinen","doi":"10.1109/OJVT.2026.3656394","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3656394","url":null,"abstract":"The rise of data-intensive applications in Fifth-Generation (5G) mobile networks demands that next-generation mobile systems deliver seamless, high-bandwidth, and immersive services with improved quality of service. To address these challenges, the use of Integrated Access and Backhaul (IAB) nodes operating over millimeter-wave frequency bands onboard Unmanned Aerial Vehicle (UAV) presents a promising solution. The UAV-mounted IAB network has the potential to enhance line-of-sight conditions to the donor Base Station (BS) via the backhaul link, enabling temporary high data rates in mission-critical and emergency response communication scenarios that require a rapid deployment of new network elements for boosting cellular coverage. This article proposed a framework that integrates terrestrial IAB nodes, non-terrestrial UAV-mounted IAB nodes, and terrestrial donor BS, operating in a dense urban Manhattan-like environment. The research work primarily focuses on how variations in UAV-mounted IAB altitudes, donor BS down-tilt angle, and IAB antenna configuration influence the downlink end-to-end (E2E) spectral efficiency performance of mobile users. Simulation results demonstrate that significant performance gains can be achieved when non-terrestrial IAB nodes are deployed at suitable altitudes when equipped with appropriate antenna configurations. These improvements are further improved when the donor BS employs properly adjusted down-tilt angles, enabling the hybrid terrestrial-aerial IAB mobile network to operate more efficiently and deliver enhanced E2E performances.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"676-690"},"PeriodicalIF":4.8,"publicationDate":"2026-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11359721","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146223751","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 : 2026-01-20DOI: 10.1109/OJVT.2026.3656339
Zoltán Márton;István Szalay;Dénes Fodor
In this paper, we propose a novel signal processing method for road surface roughness classification exclusively from wheel speed sensor signals. Road surface quality has a significant impact on fuel consumption and driving safety. Traditionally, it has been measured using specially equipped vehicles and, more recently, shared via cloud-based infrastructure; however, such data can be unavailable or quickly become outdated, making onboard solutions essential. We analyzed a large wheel speed sensor dataset from various test maneuvers to determine how road surface roughness influences spectral characteristics under different conditions, including changes in speed, tire pressure, payload, and tire type. The proposed road surface roughness classifier uses only wheel speed sensor signals. It selects signal segments appropriate for processing based on driving conditions and computes their order spectra. The number and relative power of the spectral peaks within the identified interval of interest of the order spectrum are related to road surface roughness. The implemented classifier is capable of distinguishing between rough and smooth road surfaces based on the number of peaks in the interval of interest. The overall accuracy of the implemented road surface roughness classifier was $87.4 ,%$.
{"title":"Wheel-Speed-Sensor-Based Spectral Classifier for Road Surface Roughness","authors":"Zoltán Márton;István Szalay;Dénes Fodor","doi":"10.1109/OJVT.2026.3656339","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3656339","url":null,"abstract":"In this paper, we propose a novel signal processing method for road surface roughness classification exclusively from wheel speed sensor signals. Road surface quality has a significant impact on fuel consumption and driving safety. Traditionally, it has been measured using specially equipped vehicles and, more recently, shared via cloud-based infrastructure; however, such data can be unavailable or quickly become outdated, making onboard solutions essential. We analyzed a large wheel speed sensor dataset from various test maneuvers to determine how road surface roughness influences spectral characteristics under different conditions, including changes in speed, tire pressure, payload, and tire type. The proposed road surface roughness classifier uses only wheel speed sensor signals. It selects signal segments appropriate for processing based on driving conditions and computes their order spectra. The number and relative power of the spectral peaks within the identified interval of interest of the order spectrum are related to road surface roughness. The implemented classifier is capable of distinguishing between rough and smooth road surfaces based on the number of peaks in the interval of interest. The overall accuracy of the implemented road surface roughness classifier was <inline-formula><tex-math>$87.4 ,%$</tex-math></inline-formula>.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"829-843"},"PeriodicalIF":4.8,"publicationDate":"2026-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11359493","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147362559","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 : 2026-01-20DOI: 10.1109/OJVT.2026.3655621
Zekun Hong;Shinya Sugiura;Chao Xu;Lajos Hanzo
We conceive a novel channel estimation and data detection scheme for OTFS-modulated faster-than-Nyquist (FTN) transmission over doubly selective fading channels, aiming for enhancing the spectral efficiency and Doppler resilience. The delay-Doppler (DD) domain's input-output relationship of OTFS-FTN signaling is derived by employing a root-raised cosine (RRC) shaping filter. More specifically, we design our DD-domain channel estimator for FTN-based pilot transmission, where the pilot symbol interval is lower than that defined by the classic Nyquist criterion. Moreover, we propose a reduced-complexity linear minimum mean square error equalizer, supporting noise whitening, where the FTN-induced inter-symbol interference (ISI) matrix is approximated by a sparse one. Our performance results demonstrate that the proposed OTFS-FTN scheme is capable of enhancing the achievable information rate, while attaining a comparable BER performance to both that of its Nyquist-based OTFS counterpart and to other FTN transmission schemes, which employ the same RRC shaping filter.
{"title":"Delay-Doppler-Domain Channel Estimation and Reduced-Complexity Detection of Faster-Than-Nyquist Signaling Aided OTFS","authors":"Zekun Hong;Shinya Sugiura;Chao Xu;Lajos Hanzo","doi":"10.1109/OJVT.2026.3655621","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3655621","url":null,"abstract":"We conceive a novel channel estimation and data detection scheme for OTFS-modulated faster-than-Nyquist (FTN) transmission over doubly selective fading channels, aiming for enhancing the spectral efficiency and Doppler resilience. The delay-Doppler (DD) domain's input-output relationship of OTFS-FTN signaling is derived by employing a root-raised cosine (RRC) shaping filter. More specifically, we design our DD-domain channel estimator for FTN-based pilot transmission, where the pilot symbol interval is lower than that defined by the classic Nyquist criterion. Moreover, we propose a reduced-complexity linear minimum mean square error equalizer, supporting noise whitening, where the FTN-induced inter-symbol interference (ISI) matrix is approximated by a sparse one. Our performance results demonstrate that the proposed OTFS-FTN scheme is capable of enhancing the achievable information rate, while attaining a comparable BER performance to both that of its Nyquist-based OTFS counterpart and to other FTN transmission schemes, which employ the same RRC shaping filter.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"659-675"},"PeriodicalIF":4.8,"publicationDate":"2026-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11359468","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146175614","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 : 2026-01-16DOI: 10.1109/OJVT.2026.3651868
Edward Au
{"title":"Editor-in-Chief's Messages With Gratitude and Pride: A Year of Growth and Shared Excellence","authors":"Edward Au","doi":"10.1109/OJVT.2026.3651868","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3651868","url":null,"abstract":"","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"i-i"},"PeriodicalIF":4.8,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11356006","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145982307","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 design of automated driving systems is of growing industry and societal interest. Perception is a critical technology for these systems, which allows a vehicle to discern the surrounding environment. Perception systems in automated vehicles frequently use machine vision algorithms; however, the performance of a machine vision algorithm critically depends on the quality of the data provided. Quantifying the ‘quality’ of image data is therefore potentially a useful tool in understanding and predicting the performance of a machine vision system. This study uses the Shannon Information Capacity, a metric based on information theory, to evaluate the impact of image quality on a perception algorithm. In this preliminary study, a set of synthetic objects are arranged to create a novel simulated test chart. The chart contains standard machine vision objects of interest (people and cars) as well as a slanted edge, which is used to calculate image quality metrics. The chart is degraded using varying levels of contrast and blur to simulate different real-world operating conditions. Object detection performance is then evaluated using a range of deep learning-based detection algorithms, with different architectures. The results indicate that Shannon Information Capacity has the potential to predict machine vision performance across multiple model architectures and object types. For example, the results for all the models show that accuracy remains relatively constant above an SIC value of 0.25 b/p. Results indicate that for YOLOv10 m SIC has mutual information value with detection accuracy of 1.66 bits while MTF50 has a score of 0.4945 bits. This study is the first to show the correlation between SIC and machine vision performance. While other metrics have been previously shown to have some correlation with machine vision, the correlation shown by SIC is much stronger. The findings presented may be of use to designers of autonomous driving systems and automotive camera manufacturers.
自动驾驶系统的设计越来越受到工业界和社会的关注。感知是这些系统的关键技术，它允许车辆识别周围环境。自动驾驶汽车中的感知系统经常使用机器视觉算法；然而，机器视觉算法的性能很大程度上取决于所提供数据的质量。因此，量化图像数据的“质量”可能是理解和预测机器视觉系统性能的有用工具。本研究使用香农信息容量，一个基于信息论的度量，来评估图像质量对感知算法的影响。在这个初步的研究中，我们利用一组合成对象来创建一个新的模拟测试图。该图表包含感兴趣的标准机器视觉对象（人和汽车）以及用于计算图像质量指标的倾斜边缘。使用不同水平的对比度和模糊来模拟不同的现实世界操作条件，图表被降级。然后使用一系列具有不同架构的基于深度学习的检测算法来评估目标检测性能。结果表明，香农信息能力具有跨多种模型架构和对象类型预测机器视觉性能的潜力。例如，所有模型的结果表明，在SIC值为0.25 b/p以上，精度保持相对恒定。结果表明，YOLOv10 m SIC具有互信息值，检测精度为1.66 bits， MTF50得分为0.4945 bits。这项研究首次展示了SIC与机器视觉性能之间的相关性。虽然其他指标之前已经被证明与机器视觉有一定的相关性，但SIC显示的相关性要强得多。这些发现可能会对自动驾驶系统的设计者和汽车摄像头制造商有所帮助。
{"title":"Information Capacity as a Predictor of Perception Performance","authors":"Diarmaid Geever;Tim Brophy;Dara Molloy;Enda Ward;Roshan George;Norman Koren;Martin Glavin;Edward Jones;Brian Deegan","doi":"10.1109/OJVT.2026.3655075","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3655075","url":null,"abstract":"The design of automated driving systems is of growing industry and societal interest. Perception is a critical technology for these systems, which allows a vehicle to discern the surrounding environment. Perception systems in automated vehicles frequently use machine vision algorithms; however, the performance of a machine vision algorithm critically depends on the quality of the data provided. Quantifying the ‘quality’ of image data is therefore potentially a useful tool in understanding and predicting the performance of a machine vision system. This study uses the Shannon Information Capacity, a metric based on information theory, to evaluate the impact of image quality on a perception algorithm. In this preliminary study, a set of synthetic objects are arranged to create a novel simulated test chart. The chart contains standard machine vision objects of interest (people and cars) as well as a slanted edge, which is used to calculate image quality metrics. The chart is degraded using varying levels of contrast and blur to simulate different real-world operating conditions. Object detection performance is then evaluated using a range of deep learning-based detection algorithms, with different architectures. The results indicate that Shannon Information Capacity has the potential to predict machine vision performance across multiple model architectures and object types. For example, the results for all the models show that accuracy remains relatively constant above an SIC value of 0.25 b/p. Results indicate that for YOLOv10 m SIC has mutual information value with detection accuracy of 1.66 bits while MTF50 has a score of 0.4945 bits. This study is the first to show the correlation between SIC and machine vision performance. While other metrics have been previously shown to have some correlation with machine vision, the correlation shown by SIC is much stronger. The findings presented may be of use to designers of autonomous driving systems and automotive camera manufacturers.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"708-722"},"PeriodicalIF":4.8,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11355802","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146223753","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}
Machine learning (ML) and deep learning (DL) have become essential tools in lithium-ion battery research, particularly for estimating the State of Health (SOH). However, conventional SOH estimation methods often rely on repeated charge/discharge cycles under strictly controlled laboratory conditions, limiting their applicability in real world scenarios. In this study, we present a comprehensive lithium-ion battery dataset developed by our team to support data driven approaches for battery diagnostics and predictive modeling. The dataset comprises nineteen lithium iron phosphate (LFP) cells with cycle lifetimes ranging from 500 to 2600 cycles and reflects realistic usage conditions, including non constant discharge currents and tests conducted at $25,^circ text{C}$, $35,^circ text{C}$, and $45,^circ text{C}$. To demonstrate the utility of this dataset, we used a brain inspired Spiking Neural Network (SNN) referred to as SpikeSOH, a neuromorphic model that uses sparse, time coded spikes to mimic biological neurons. This approach provides temporal precision while reducing energy consumption. Our results show that the SNN-based model achieves an average Mean Absolute Error (MAE) of 4.5%, while also outperforming conventional deep learning models in computational efficiency, with an average inference time of 3.55 $mu mathrm{s}$ and an average energy consumption of 0.36 mJ. These characteristics make the model particularly suitable for integration into energy constrained battery management systems. By providing a realistic, high quality dataset and demonstrating the advantages of energy efficient neuromorphic models, this work advances accurate and scalable SOH estimation methods, helping safer and more reliable deployment of lithium-ion batteries in both first life and second life applications.
{"title":"Spiking Neural Networks for Accurate and Efficient State of Health Estimation of Lithium-Ion Batteries Across Varying Temperatures","authors":"Slimane Arbaoui;Tedjani Mesbahi;Théo Heitzmann;Marwa Zitouni;Amel Hidouri;Lakhdar Mamouri;Ali Ayadi;Ahmed Samet;Romuald Boné","doi":"10.1109/OJVT.2026.3653419","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3653419","url":null,"abstract":"Machine learning (ML) and deep learning (DL) have become essential tools in lithium-ion battery research, particularly for estimating the State of Health (SOH). However, conventional SOH estimation methods often rely on repeated charge/discharge cycles under strictly controlled laboratory conditions, limiting their applicability in real world scenarios. In this study, we present a comprehensive lithium-ion battery dataset developed by our team to support data driven approaches for battery diagnostics and predictive modeling. The dataset comprises nineteen lithium iron phosphate (LFP) cells with cycle lifetimes ranging from 500 to 2600 cycles and reflects realistic usage conditions, including non constant discharge currents and tests conducted at <inline-formula><tex-math>$25,^circ text{C}$</tex-math></inline-formula>, <inline-formula><tex-math>$35,^circ text{C}$</tex-math></inline-formula>, and <inline-formula><tex-math>$45,^circ text{C}$</tex-math></inline-formula>. To demonstrate the utility of this dataset, we used a brain inspired Spiking Neural Network (SNN) referred to as SpikeSOH, a neuromorphic model that uses sparse, time coded spikes to mimic biological neurons. This approach provides temporal precision while reducing energy consumption. Our results show that the SNN-based model achieves an average Mean Absolute Error (MAE) of 4.5%, while also outperforming conventional deep learning models in computational efficiency, with an average inference time of 3.55 <inline-formula><tex-math>$mu mathrm{s}$</tex-math></inline-formula> and an average energy consumption of 0.36 mJ. These characteristics make the model particularly suitable for integration into energy constrained battery management systems. By providing a realistic, high quality dataset and demonstrating the advantages of energy efficient neuromorphic models, this work advances accurate and scalable SOH estimation methods, helping safer and more reliable deployment of lithium-ion batteries in both first life and second life applications.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"510-522"},"PeriodicalIF":4.8,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11347466","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146082179","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 : 2026-01-14DOI: 10.1109/OJVT.2026.3653504
O. Tokluoglu;A. Cicek;E. Cavus;E. Bedeer;H. Yanikomeroglu
This paper presents a deep learning-based detector for faster-than-Nyquist (FTN) signaling that leverages a Gated Recurrent Unit (GRU) architecture optimized using the Nesterov-accelerated Adaptive Moment Estimation (NADAM) algorithm. Compared with Long Short-Term Memory (LSTM) networks commonly employed in similar detection tasks, GRUs offer improved computational efficiency, while NADAM contributes to stable and effective convergence in non-convex optimization settings. Rather than relying on generic neural models, the proposed design explicitly aligns the GRU input structure with the one-sided inter-symbol interference (ISI) span of FTN signaling, enabling the network to learn interference patterns efficiently without incurring unnecessary complexity. This structured integration results in reduced computational burden and enhanced convergence behavior. Simulation results demonstrate that the NADAM-optimized GRU achieves bit error rate (BER) performance close to the optimal BCJR algorithm for $tau geq 0.7$, while offering superior computational efficiency compared with conventional deep learning-based detectors. A detailed complexity comparison with the M-BCJR algorithm shows that the proposed approach reduces hardware resource usage—measured in Look-Up Tables (LUTs)—by up to 76% while maintaining comparable BER performance in the same $tau$ regime. Additional evaluations further highlight its robustness, demonstrating reliable performance in quasi-static multipath Rayleigh fading channels and strong compatibility with LDPC-coded FTN transmission. These results collectively underscore the practicality and efficiency of the proposed GRU-based FTN detector.
本文提出了一种基于深度学习的检测器，用于比奈奎斯特（FTN）信号更快的信号，该检测器利用了使用nesterov加速自适应矩估计（NADAM）算法优化的门控循环单元（GRU）架构。与用于类似检测任务的长短期记忆（LSTM）网络相比，gru提供了更高的计算效率，而NADAM有助于在非凸优化设置下稳定有效的收敛。该设计不依赖于通用神经模型，而是明确地将GRU输入结构与FTN信令的单侧符号间干扰（ISI）跨度对齐，使网络能够有效地学习干扰模式，而不会产生不必要的复杂性。这种结构化集成减少了计算负担，增强了收敛性。仿真结果表明，经过nadam优化的GRU的误码率（BER）性能接近于$tau geq 0.7$的最优BCJR算法，同时与传统的基于深度学习的检测器相比，具有更高的计算效率。与M-BCJR算法的详细复杂性比较表明，所提出的方法减少了硬件资源的使用（以查找表（LUTs）衡量）高达76% while maintaining comparable BER performance in the same $tau$ regime. Additional evaluations further highlight its robustness, demonstrating reliable performance in quasi-static multipath Rayleigh fading channels and strong compatibility with LDPC-coded FTN transmission. These results collectively underscore the practicality and efficiency of the proposed GRU-based FTN detector.
{"title":"GRU-Based Sequence Detection for Faster-than-Nyquist Signaling","authors":"O. Tokluoglu;A. Cicek;E. Cavus;E. Bedeer;H. Yanikomeroglu","doi":"10.1109/OJVT.2026.3653504","DOIUrl":"https://doi.org/10.1109/OJVT.2026.3653504","url":null,"abstract":"This paper presents a deep learning-based detector for faster-than-Nyquist (FTN) signaling that leverages a Gated Recurrent Unit (GRU) architecture optimized using the Nesterov-accelerated Adaptive Moment Estimation (NADAM) algorithm. Compared with Long Short-Term Memory (LSTM) networks commonly employed in similar detection tasks, GRUs offer improved computational efficiency, while NADAM contributes to stable and effective convergence in non-convex optimization settings. Rather than relying on generic neural models, the proposed design explicitly aligns the GRU input structure with the one-sided inter-symbol interference (ISI) span of FTN signaling, enabling the network to learn interference patterns efficiently without incurring unnecessary complexity. This structured integration results in reduced computational burden and enhanced convergence behavior. Simulation results demonstrate that the NADAM-optimized GRU achieves bit error rate (BER) performance close to the optimal BCJR algorithm for <inline-formula><tex-math>$tau geq 0.7$</tex-math></inline-formula>, while offering superior computational efficiency compared with conventional deep learning-based detectors. A detailed complexity comparison with the M-BCJR algorithm shows that the proposed approach reduces hardware resource usage—measured in Look-Up Tables (LUTs)—by up to 76% while maintaining comparable BER performance in the same <inline-formula><tex-math>$tau$</tex-math></inline-formula> regime. Additional evaluations further highlight its robustness, demonstrating reliable performance in quasi-static multipath Rayleigh fading channels and strong compatibility with LDPC-coded FTN transmission. These results collectively underscore the practicality and efficiency of the proposed GRU-based FTN detector.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"7 ","pages":"565-581"},"PeriodicalIF":4.8,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11352857","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146175647","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：