Pub Date : 2026-01-01DOI: 10.1109/ACCESS.2025.3650116
Guoqing Zhang;Yang Liu;Zhiliang Chen;Yichun Wang
Reliable fault diagnosis of DC-DC converters is critical for ensuring the safety and operational continuity of electric vehicles (EVs) and industrial power systems. In practical applications, early detection of progressive degradation can prevent catastrophic failures and significantly reduce maintenance costs. However, developing robust diagnostic models is hindered by sensor noise, environmental disturbances, and the scarcity of labeled fault data—particularly for subtle degradation scenarios where real-world collection is expensive. To address these issues, this paper proposes a transfer learning framework based on Inception-Residual Neural Network (ResNet) + One-Dimensional Convolutional Neural Network (1D-CNN), aiming to enhance diagnostic performance under small-sample conditions. The proposed network extracts discriminative features from simulated voltage signals (source domain) and transfers this knowledge to real-world experimental data (target domain). To reduce domain discrepancies, a Whitening and Coloring Covariance Alignment (WCA) is employed for global feature alignment, while a novel Classwise WCA strategy further refines the alignment at the category level to preserve fault-specific structures. Additionally, the classifier is fine-tuned using L2 Starting Point (L2-SP) regularization, which constrains parameter shift to prevent overfitting under limited supervision. The method is non-intrusive and practical, operating efficiently using only the output voltage ripple. Experimental results validate the effectiveness of the proposed approach, demonstrating a diagnostic accuracy of 91.6% even with limited samples, significantly outperforming conventional transfer learning methods in cross-domain fault identification tasks.
{"title":"A Novel Transfer Learning Strategy Based on Inception-ResNet+ 1D-CNN for DC–DC Converter Degradation Fault Diagnosis","authors":"Guoqing Zhang;Yang Liu;Zhiliang Chen;Yichun Wang","doi":"10.1109/ACCESS.2025.3650116","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3650116","url":null,"abstract":"Reliable fault diagnosis of DC-DC converters is critical for ensuring the safety and operational continuity of electric vehicles (EVs) and industrial power systems. In practical applications, early detection of progressive degradation can prevent catastrophic failures and significantly reduce maintenance costs. However, developing robust diagnostic models is hindered by sensor noise, environmental disturbances, and the scarcity of labeled fault data—particularly for subtle degradation scenarios where real-world collection is expensive. To address these issues, this paper proposes a transfer learning framework based on Inception-Residual Neural Network (ResNet) + One-Dimensional Convolutional Neural Network (1D-CNN), aiming to enhance diagnostic performance under small-sample conditions. The proposed network extracts discriminative features from simulated voltage signals (source domain) and transfers this knowledge to real-world experimental data (target domain). To reduce domain discrepancies, a Whitening and Coloring Covariance Alignment (WCA) is employed for global feature alignment, while a novel Classwise WCA strategy further refines the alignment at the category level to preserve fault-specific structures. Additionally, the classifier is fine-tuned using L2 Starting Point (L2-SP) regularization, which constrains parameter shift to prevent overfitting under limited supervision. The method is non-intrusive and practical, operating efficiently using only the output voltage ripple. Experimental results validate the effectiveness of the proposed approach, demonstrating a diagnostic accuracy of 91.6% even with limited samples, significantly outperforming conventional transfer learning methods in cross-domain fault identification tasks.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"10667-10680"},"PeriodicalIF":3.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11321294","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146026406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study proposes an improved object detection model, Agri-YOLO, based on the YOLOv8n baseline model, addressing the issues of insufficient detection accuracy and high model deployment costs for various types and scales of obstacles in agricultural fields. Three core optimization strategies are implemented to enhance model performance: replacing traditional convolution with the Wavelet Transform Convolution module to improve multi-scale feature perception with only a slight increase in parameters; utilizing the Wise-IoU loss function to optimize bounding box regression, enhancing the localization accuracy of irregular obstacles and effectively improving the convergence speed and regression accuracy of the loss function; and integrating the Dynamic Upsample module to reduce computational load while ensuring detection accuracy of agricultural obstacles, thereby improving feature recovery accuracy. Experimental results demonstrate that Agri-YOLO significantly outperforms baseline algorithms such as Faster R-CNN, SSD, and YOLOv8n in terms of precision, recall, and mAP50 metrics, with improvements of 0.4, 1.7, and 1.2% in accuracy, recall, and mAP0.5, respectively, while also enhancing model robustness and stability. This research provides an efficient technical solution for detecting obstacles in agricultural fields.
{"title":"Agri-YOLO: An Improved YOLOv8 Algorithm for Farmland Obstacles Detection","authors":"Xiang Gan;Mengjie Xing;Shukun Cao;Wenhao Zhang;Yu Wang;Li Zeng;Lei Xu","doi":"10.1109/ACCESS.2025.3650351","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3650351","url":null,"abstract":"This study proposes an improved object detection model, Agri-YOLO, based on the YOLOv8n baseline model, addressing the issues of insufficient detection accuracy and high model deployment costs for various types and scales of obstacles in agricultural fields. Three core optimization strategies are implemented to enhance model performance: replacing traditional convolution with the Wavelet Transform Convolution module to improve multi-scale feature perception with only a slight increase in parameters; utilizing the Wise-IoU loss function to optimize bounding box regression, enhancing the localization accuracy of irregular obstacles and effectively improving the convergence speed and regression accuracy of the loss function; and integrating the Dynamic Upsample module to reduce computational load while ensuring detection accuracy of agricultural obstacles, thereby improving feature recovery accuracy. Experimental results demonstrate that Agri-YOLO significantly outperforms baseline algorithms such as Faster R-CNN, SSD, and YOLOv8n in terms of precision, recall, and mAP50 metrics, with improvements of 0.4, 1.7, and 1.2% in accuracy, recall, and mAP0.5, respectively, while also enhancing model robustness and stability. This research provides an efficient technical solution for detecting obstacles in agricultural fields.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"10828-10840"},"PeriodicalIF":3.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11321291","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146026374","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-01DOI: 10.1109/ACCESS.2025.3650213
Qidong Yan;Chenglin Jiang;Yingjie Li;Ning Ma
Large language models (LLMs) have achieved remarkable progress in natural language processing, yet their ability to perform complex logical reasoning remains limited. Existing approaches such as prompting, retrieval-augmented generation, and parameter-efficient fine-tuning (PEFT) provide partial improvements but often suffer from prompt sensitivity, semantic compression, or additional computational cost. In this work, we propose a Residual Feature Enhancement (RFE) module, a lightweight architectural component designed to strengthen reasoning ability while maintaining computational efficiency. RFE integrates a dimension-preserving linear transformation, SwiGLU nonlinear activation, and residual connections to enrich attention outputs without altering the backbone structure. We conducted comprehensive experiments across six reasoning and comprehension benchmarks—LogiQA, ReClor, LogiQA2.0, GSM8K, HellaSwag, and MBPP—covering deductive reasoning, standardized test comprehension, commonsense inference, and program synthesis. Results demonstrate that ChatGLM4-9B augmented with RFE consistently achieves superior performance compared with both adapter-based methods and larger-scale baselines. Specifically, ChatGLM4-9B+ RFE attains 68.20% on LogiQA, 82.00% on ReClor, 79.74% on LogiQA 2.0, 95.68% on GSM8K, 72.42% on HellaSwag, and 56.82% on MBPP, all of which surpass the Adapter mechanism (67.68%, 81.15%, 78.52%, 94.47%, 66.85%, 55.02%) and show clear advantages over open-source baselines such as Qwen1.5-MoE-A2.7B, Llama3.1-8B, and DeepSeek distilled models. Ablation studies further confirm that removing RFE leads to performance degradation of up to 3.84 percentage points, and convergence analysis shows improved stability and faster training.
{"title":"Residual Feature Enhancement for Large Language Models: Methodology and Applications","authors":"Qidong Yan;Chenglin Jiang;Yingjie Li;Ning Ma","doi":"10.1109/ACCESS.2025.3650213","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3650213","url":null,"abstract":"Large language models (LLMs) have achieved remarkable progress in natural language processing, yet their ability to perform complex logical reasoning remains limited. Existing approaches such as prompting, retrieval-augmented generation, and parameter-efficient fine-tuning (PEFT) provide partial improvements but often suffer from prompt sensitivity, semantic compression, or additional computational cost. In this work, we propose a Residual Feature Enhancement (RFE) module, a lightweight architectural component designed to strengthen reasoning ability while maintaining computational efficiency. RFE integrates a dimension-preserving linear transformation, SwiGLU nonlinear activation, and residual connections to enrich attention outputs without altering the backbone structure. We conducted comprehensive experiments across six reasoning and comprehension benchmarks—LogiQA, ReClor, LogiQA2.0, GSM8K, HellaSwag, and MBPP—covering deductive reasoning, standardized test comprehension, commonsense inference, and program synthesis. Results demonstrate that ChatGLM4-9B augmented with RFE consistently achieves superior performance compared with both adapter-based methods and larger-scale baselines. Specifically, ChatGLM4-9B+ RFE attains 68.20% on LogiQA, 82.00% on ReClor, 79.74% on LogiQA 2.0, 95.68% on GSM8K, 72.42% on HellaSwag, and 56.82% on MBPP, all of which surpass the Adapter mechanism (67.68%, 81.15%, 78.52%, 94.47%, 66.85%, 55.02%) and show clear advantages over open-source baselines such as Qwen1.5-MoE-A2.7B, Llama3.1-8B, and DeepSeek distilled models. Ablation studies further confirm that removing RFE leads to performance degradation of up to 3.84 percentage points, and convergence analysis shows improved stability and faster training.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"4052-4062"},"PeriodicalIF":3.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11321298","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145929535","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-01DOI: 10.1109/ACCESS.2025.3650467
Damian Kedziora;Arkadiusz Jurczuk
This study examines how communication structures can be governed to improve outcomes in distributed project teams implementing Robotic Process Automation (RPA). Using a qualitative single-case study at SSAB, we analyse semi-structured interviews, project artefacts, and internal communications, iterating abductively between data and theory. Guided by Stewardship Theory and project governance research, we map decision rights, accountability, and information flows across locations and functions. Findings show that adaptive communication networks anchored in explicit governance mechanisms, i.e. clear role charters, cadence calendars, escalation paths, and gated decision points, reduce ambiguity and coordination loss. Informal practices, including cross-site champions and community-of-practice touchpoints, complement formal protocols by brokering knowledge and sustaining shared purpose. The interplay between formal and informal structures improved timeliness of information, alignment on deliverables, and proactive risk handling in geographically dispersed settings. We propose a practical design for communication governance that specifies who convenes whom, on what cadence, with what artefacts, and how exceptions escalate. The study extends Stewardship Theory to automation-oriented, distributed projects by showing how trust-based, collective-interest framing can coexist with lightweight controls. Implications include a diagnostic for assessing communication gaps and actionable guidance for configuring roles, routines, and artefacts when scaling RPA initiatives across sites.
{"title":"Governing Communication Structure Across Distributed Teams at Projects Implementing Process Automation Software","authors":"Damian Kedziora;Arkadiusz Jurczuk","doi":"10.1109/ACCESS.2025.3650467","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3650467","url":null,"abstract":"This study examines how communication structures can be governed to improve outcomes in distributed project teams implementing Robotic Process Automation (RPA). Using a qualitative single-case study at SSAB, we analyse semi-structured interviews, project artefacts, and internal communications, iterating abductively between data and theory. Guided by Stewardship Theory and project governance research, we map decision rights, accountability, and information flows across locations and functions. Findings show that adaptive communication networks anchored in explicit governance mechanisms, i.e. clear role charters, cadence calendars, escalation paths, and gated decision points, reduce ambiguity and coordination loss. Informal practices, including cross-site champions and community-of-practice touchpoints, complement formal protocols by brokering knowledge and sustaining shared purpose. The interplay between formal and informal structures improved timeliness of information, alignment on deliverables, and proactive risk handling in geographically dispersed settings. We propose a practical design for communication governance that specifies who convenes whom, on what cadence, with what artefacts, and how exceptions escalate. The study extends Stewardship Theory to automation-oriented, distributed projects by showing how trust-based, collective-interest framing can coexist with lightweight controls. Implications include a diagnostic for assessing communication gaps and actionable guidance for configuring roles, routines, and artefacts when scaling RPA initiatives across sites.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"7681-7698"},"PeriodicalIF":3.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11322684","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145982129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-31DOI: 10.1109/ACCESS.2025.3650056
Satirtha Paul Shyam;Shaikh Anowarul Fattah;Mohammad Saquib
Medical image segmentation often involves inherent uncertainty due to inter observer variability. In this case, a single deterministic mask obtained by conventional segmentation networks, such as U-Net, cannot capture the distribution of plausible expert annotations, risking missed clinically relevant variants. In order to enable uncertainty quantification and reflect inter expert variability, probabilistic models like Probabilistic U-Net are used, which perform aleatoric or ambiguous segmentation where a latent space is sampled to generate multiple segmentation masks. However, the common use of a conditioned unimodal posterior in these models fails to represent true multimodality, leading to mode bias and limited diversity. To address these limitations, a multi-level Probabilistic U-Net augmented with normalizing flows is proposed to enhance the expressiveness of the latent distribution. The multi-level design induces multiple latent distributions in separate levels of U-Net, enabling more diverse sampling, while the flow module transforms the posterior to add data required modes and expand representational capacity, thereby enriching the expressiveness of the distributions. The proposed flow incorporated multi-level network enables a more flexible and powerful distribution, thereby enhancing the model’s ability to generate high fidelity segmentation masks. Extensive experiments on some publicly available datasets with multiple expert annotations per image demonstrate that the proposed model reduces generalized energy distance (GED), preserves clinically meaningful diversity and sharpens boundary fidelity, with latent grid analyses indicating fuller mode coverage and fewer artifacts. Collectively, these results indicate that the proposed framework advances accuracy, robustness, and clinical reliability for aleatoric, uncertainty aware medical image segmentation.
{"title":"A Multi-Level Probabilistic Deep Learning Network Augmented With Normalizing Flow for Ambiguous Medical Image Segmentation","authors":"Satirtha Paul Shyam;Shaikh Anowarul Fattah;Mohammad Saquib","doi":"10.1109/ACCESS.2025.3650056","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3650056","url":null,"abstract":"Medical image segmentation often involves inherent uncertainty due to inter observer variability. In this case, a single deterministic mask obtained by conventional segmentation networks, such as U-Net, cannot capture the distribution of plausible expert annotations, risking missed clinically relevant variants. In order to enable uncertainty quantification and reflect inter expert variability, probabilistic models like Probabilistic U-Net are used, which perform aleatoric or ambiguous segmentation where a latent space is sampled to generate multiple segmentation masks. However, the common use of a conditioned unimodal posterior in these models fails to represent true multimodality, leading to mode bias and limited diversity. To address these limitations, a multi-level Probabilistic U-Net augmented with normalizing flows is proposed to enhance the expressiveness of the latent distribution. The multi-level design induces multiple latent distributions in separate levels of U-Net, enabling more diverse sampling, while the flow module transforms the posterior to add data required modes and expand representational capacity, thereby enriching the expressiveness of the distributions. The proposed flow incorporated multi-level network enables a more flexible and powerful distribution, thereby enhancing the model’s ability to generate high fidelity segmentation masks. Extensive experiments on some publicly available datasets with multiple expert annotations per image demonstrate that the proposed model reduces generalized energy distance (GED), preserves clinically meaningful diversity and sharpens boundary fidelity, with latent grid analyses indicating fuller mode coverage and fewer artifacts. Collectively, these results indicate that the proposed framework advances accuracy, robustness, and clinical reliability for aleatoric, uncertainty aware medical image segmentation.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"4063-4079"},"PeriodicalIF":3.6,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11321095","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145929343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-31DOI: 10.1109/ACCESS.2025.3649779
Chien-Wei Hu;Yi-Hsuan Liao;Hewijin Christine Jiau
Enhancing programming skills is essential for developers to keep pace with technological advancements and to maintain effective participation in software development practices. Game-based programming platforms have been widely adopted to promote learner engagement and skill acquisition. However, without structured guidance, programmers may adopt ineffective strategies, leading to stagnation and wasted effort. This paper investigates the programming skills developed through ELOP, a competitive game-based training platform that has accumulated longitudinal programming data from hundreds of users. A mixed-methods analysis reveals that while ELOP fosters iterative strategy refinement, key skills such as writing well-documented code and refactoring maintainable programs remain difficult for many learners to become proficient. To address these challenges, we present GPST (Game-based Programming Skill Trainer), an extended platform that augments ELOP with instructional features including automated code smell detection, comment quality guidance, and targeted training materials. GPST aims to support learners in developing clean, readable, and maintainable code while preserving the motivational benefits of game-based learning. Preliminary evaluation results from a small-scale pilot study (n=5) demonstrate the feasibility of GPST and suggest positive learning outcomes, while indicating directions for larger future deployments.
{"title":"From Strategy to Structure: Guiding Code Quality With GPST in Game-Based Programming Environments","authors":"Chien-Wei Hu;Yi-Hsuan Liao;Hewijin Christine Jiau","doi":"10.1109/ACCESS.2025.3649779","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3649779","url":null,"abstract":"Enhancing programming skills is essential for developers to keep pace with technological advancements and to maintain effective participation in software development practices. Game-based programming platforms have been widely adopted to promote learner engagement and skill acquisition. However, without structured guidance, programmers may adopt ineffective strategies, leading to stagnation and wasted effort. This paper investigates the programming skills developed through ELOP, a competitive game-based training platform that has accumulated longitudinal programming data from hundreds of users. A mixed-methods analysis reveals that while ELOP fosters iterative strategy refinement, key skills such as writing well-documented code and refactoring maintainable programs remain difficult for many learners to become proficient. To address these challenges, we present GPST (Game-based Programming Skill Trainer), an extended platform that augments ELOP with instructional features including automated code smell detection, comment quality guidance, and targeted training materials. GPST aims to support learners in developing clean, readable, and maintainable code while preserving the motivational benefits of game-based learning. Preliminary evaluation results from a small-scale pilot study (n=5) demonstrate the feasibility of GPST and suggest positive learning outcomes, while indicating directions for larger future deployments.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"4152-4161"},"PeriodicalIF":3.6,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11320304","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145929298","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-31DOI: 10.1109/ACCESS.2025.3649777
Dohoon Kim
With the advent of the NewSpace era, space-based systems are facing complex, multi-vector-based cyber threats. Accordingly, a lifecycle-oriented approach to internalizing security becomes essential., and the concept of a Space Risk Management Framework (S-RMF), similar to the existing Risk Management Framework (RMF) system in the defense field, is required for space cybersecurity. Based on Model-Based Security Engineering (MBSE), this study references MITRE ATT&CK and Security and Privacy Architecture Through Threat Assessment (SPARTA) and formalizes a Threat Assessment & Remediation Analysis (TARA) model based on threats and security controls that meet CCSDS/NIST standards. To complement the static evaluation structure of the existing TARA, we propose a dynamic risk evaluation method that considers the time-based risk change rate by applying a Stochastic Differential Equation (SDE). The derived quantitative risk is linked to the lifecycle perspective of S-RMF and enables risk evolution analysis reflecting the time lags of attack, response, and control effectiveness. This framework can strengthen security reliability by linking the threat-assessment-control-assurance steps and can serve as a standardization basis for space cybersecurity policies.
{"title":"Dynamic Threat Modeling and Risk Assessment for Space Systems","authors":"Dohoon Kim","doi":"10.1109/ACCESS.2025.3649777","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3649777","url":null,"abstract":"With the advent of the NewSpace era, space-based systems are facing complex, multi-vector-based cyber threats. Accordingly, a lifecycle-oriented approach to internalizing security becomes essential., and the concept of a Space Risk Management Framework (S-RMF), similar to the existing Risk Management Framework (RMF) system in the defense field, is required for space cybersecurity. Based on Model-Based Security Engineering (MBSE), this study references MITRE ATT&CK and Security and Privacy Architecture Through Threat Assessment (SPARTA) and formalizes a Threat Assessment & Remediation Analysis (TARA) model based on threats and security controls that meet CCSDS/NIST standards. To complement the static evaluation structure of the existing TARA, we propose a dynamic risk evaluation method that considers the time-based risk change rate by applying a Stochastic Differential Equation (SDE). The derived quantitative risk is linked to the lifecycle perspective of S-RMF and enables risk evolution analysis reflecting the time lags of attack, response, and control effectiveness. This framework can strengthen security reliability by linking the threat-assessment-control-assurance steps and can serve as a standardization basis for space cybersecurity policies.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"4080-4088"},"PeriodicalIF":3.6,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11320256","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145929413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-30DOI: 10.1109/ACCESS.2025.3649591
Yang Chen;Hanieh Ahmadi;Saba Al-Rubaye
However, traditional model-based phase-shift optimization is highly sensitive to imperfect CSI and becomes computationally prohibitive for large UPA-based RIS, while existing model-free solutions relying on single-agent DRL struggle with the exponentially growing action space. This paper presents a scalable multi-agent deep Q-network (MADQN)–based RIS controller designed for large-scale UAV–RIS systems under realistic channel dynamics. An end-to-end channel inference architecture is first introduced to mitigate CSI imperfection and reconstruct stable channel representations under UAV mobility. A multi-objective formulation is then developed to jointly optimize sum rate, energy consumption, and control latency, which is transformed into a multi-agent Markov decision process (MMDP) compatible with quantized RIS hardware. Building on this formulation, a dual-agent RIS controller is proposed, in which row and column agents cooperatively determine the quantized phase configuration of a large UPA RIS. Extensive simulations demonstrate that the proposed framework significantly outperforms benchmark schemes, showing acceptable robustness against varying Rician factor SNRs, UAV densities, and RIS sizes. These results confirm that the proposed MADQN-based controller is a promising and practical solution for scalable RIS control in large-scale multi-UAV communication systems.
{"title":"Multi-Agent Deep Reinforcement Learning-Based RIS-Aided UAV Communications","authors":"Yang Chen;Hanieh Ahmadi;Saba Al-Rubaye","doi":"10.1109/ACCESS.2025.3649591","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3649591","url":null,"abstract":"However, traditional model-based phase-shift optimization is highly sensitive to imperfect CSI and becomes computationally prohibitive for large UPA-based RIS, while existing model-free solutions relying on single-agent DRL struggle with the exponentially growing action space. This paper presents a scalable multi-agent deep Q-network (MADQN)–based RIS controller designed for large-scale UAV–RIS systems under realistic channel dynamics. An end-to-end channel inference architecture is first introduced to mitigate CSI imperfection and reconstruct stable channel representations under UAV mobility. A multi-objective formulation is then developed to jointly optimize sum rate, energy consumption, and control latency, which is transformed into a multi-agent Markov decision process (MMDP) compatible with quantized RIS hardware. Building on this formulation, a dual-agent RIS controller is proposed, in which row and column agents cooperatively determine the quantized phase configuration of a large UPA RIS. Extensive simulations demonstrate that the proposed framework significantly outperforms benchmark schemes, showing acceptable robustness against varying Rician factor SNRs, UAV densities, and RIS sizes. These results confirm that the proposed MADQN-based controller is a promising and practical solution for scalable RIS control in large-scale multi-UAV communication systems.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"1522-1536"},"PeriodicalIF":3.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11318351","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145898247","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-30DOI: 10.1109/ACCESS.2025.3649703
Jibaek Oh;Kwangphil Park;Jihoon Yoon;Junghan Kwon
Projector-based Augmented Reality (AR) provides intuitive guidance by projecting virtual content directly onto a physical workspace. However, when projected images overlap real objects, they distort the camera view, and detection accuracy drops for models like MediaPipe and YOLO. Moreover, a phenomenon known as ‘Visual Echo’, situation that system mistakes bright projected shapes as real objects, can occur. In addition, the projector-camera response is highly non-linear, which makes simple real-time correction difficult. To overcome these issues, we present a two-stage image preprocessing algorithm designed to suppress projection interference. Our method combines Color Refinement based on a Color Transformation Table and masked Lightness Compensation to effectively remove projection artifacts and enhance the visibility of physical objects. Experimental results show that the algorithm significantly reduces positional error by 70.47% and instability by 70.17% in MediaPipe hand landmark detection, while achieving 100% correct detection rate and reducing positional error by 86.32% in YOLOv8 object detection by effectively eliminating visual echoes. Furthermore, our algorithm maintains real-time performance at 27.4 FPS, making it suitable for practical applications. We successfully demonstrate the robust performance of our method through three distinct use cases: AR-based virtual ring try-on, dining etiquette education, and assembly training, highlighting its potential to enhance the reliability of projector-based AR systems across various fields.
{"title":"Visual Interference Suppression for Physical Object Detection in Projector-Based AR System","authors":"Jibaek Oh;Kwangphil Park;Jihoon Yoon;Junghan Kwon","doi":"10.1109/ACCESS.2025.3649703","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3649703","url":null,"abstract":"Projector-based Augmented Reality (AR) provides intuitive guidance by projecting virtual content directly onto a physical workspace. However, when projected images overlap real objects, they distort the camera view, and detection accuracy drops for models like MediaPipe and YOLO. Moreover, a phenomenon known as ‘Visual Echo’, situation that system mistakes bright projected shapes as real objects, can occur. In addition, the projector-camera response is highly non-linear, which makes simple real-time correction difficult. To overcome these issues, we present a two-stage image preprocessing algorithm designed to suppress projection interference. Our method combines Color Refinement based on a Color Transformation Table and masked Lightness Compensation to effectively remove projection artifacts and enhance the visibility of physical objects. Experimental results show that the algorithm significantly reduces positional error by 70.47% and instability by 70.17% in MediaPipe hand landmark detection, while achieving 100% correct detection rate and reducing positional error by 86.32% in YOLOv8 object detection by effectively eliminating visual echoes. Furthermore, our algorithm maintains real-time performance at 27.4 FPS, making it suitable for practical applications. We successfully demonstrate the robust performance of our method through three distinct use cases: AR-based virtual ring try-on, dining etiquette education, and assembly training, highlighting its potential to enhance the reliability of projector-based AR systems across various fields.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"1537-1551"},"PeriodicalIF":3.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11318862","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145898227","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-30DOI: 10.1109/ACCESS.2025.3649761
Waseem Akram;Muhayy Ud Din;Tarek Taha;Irfan Hussain
Thruster allocation is critical for the reliable operation of underwater vehicles, particularly under actuator degradation, power limitations, and thermal stress. Existing methods, such as pseudo-inverse or standard quadratic programming (QP)-based approaches, mainly minimize allocation error or energy consumption but often overlook real-time degradation and resource constraints. In this paper, we propose an adaptive fault-tolerant thrust allocation framework integrated with a PID plus Sliding Mode Control (PID+SMC) law for robust trajectory tracking. The approach leverages convex optimization to simultaneously enforce: 1) residual-driven health adaptation that down-weights degraded thrusters online; 2) power-aware allocation ensuring operation within a global energy budget; and 3) thermal-aware constraints that actively prevent overheating. A lightweight residual filter continuously updates thruster health indices, enabling rapid reallocation under faults and efficiency loss. Simulation results across nominal, power-limited, thermal-limited, faulted, and combined scenarios show that the proposed method reduces trajectory tracking error by up to 4.3% and completely eliminates power and thermal violations compared to conventional baselines. This unified framework establishes a foundation for real-time, safety-aware thruster management in marine robotics.
{"title":"Adaptive Fault-Tolerant Thrust Allocation for Underwater Vehicles With Resource Constraints","authors":"Waseem Akram;Muhayy Ud Din;Tarek Taha;Irfan Hussain","doi":"10.1109/ACCESS.2025.3649761","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3649761","url":null,"abstract":"Thruster allocation is critical for the reliable operation of underwater vehicles, particularly under actuator degradation, power limitations, and thermal stress. Existing methods, such as pseudo-inverse or standard quadratic programming (QP)-based approaches, mainly minimize allocation error or energy consumption but often overlook real-time degradation and resource constraints. In this paper, we propose an adaptive fault-tolerant thrust allocation framework integrated with a PID plus Sliding Mode Control (PID+SMC) law for robust trajectory tracking. The approach leverages convex optimization to simultaneously enforce: 1) residual-driven health adaptation that down-weights degraded thrusters online; 2) power-aware allocation ensuring operation within a global energy budget; and 3) thermal-aware constraints that actively prevent overheating. A lightweight residual filter continuously updates thruster health indices, enabling rapid reallocation under faults and efficiency loss. Simulation results across nominal, power-limited, thermal-limited, faulted, and combined scenarios show that the proposed method reduces trajectory tracking error by up to 4.3% and completely eliminates power and thermal violations compared to conventional baselines. This unified framework establishes a foundation for real-time, safety-aware thruster management in marine robotics.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"14 ","pages":"1341-1357"},"PeriodicalIF":3.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11318569","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145898248","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","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：