Computers & Electrical Engineering最新文献

{"title":"A lightweight small object detection network inspired by the visual area V2","authors":"Dandan Zhang,&nbsp;Chuan Lin,&nbsp;Yongcai Pan","doi":"10.1016/j.compeleceng.2025.110471","DOIUrl":"10.1016/j.compeleceng.2025.110471","url":null,"abstract":"<div><div>In high-resolution aerial images, due to the low feature density and small pixels of the objects to be detected, it is difficult to capture subtle details during feature extraction, which can easily lead to missed detections and false detections. To address this problem, this paper proposes a small object detection algorithm inspired by the information processing mechanism of visual area V2, in order to improve the network’s ability to extract detailed features such as the edge and direction of small objects. Inspired by the information processing mechanism of complex cells and hypercomplex cells in biological visual area V2, we designed a complex cell module (CCM) that mimics the edge-sensitive properties of complex cells and a hypercomplex cell module (HCM) that simulates the edge and direction-sensitive properties of hypercomplex cells. By simulating the sensitive characteristics of complex cells and hypercomplex cells to edge and direction information, the model’s ability to extract edge and direction features of small objects is enhanced. In addition, inspired by the bottom-up attention mechanism between visual cortexes, this paper designs a spatial enhanced attention module (SEAM) between Neck and Head, which uses shallow features to modulate deep features to retain key shallow information while focusing on small objects. The results show that on the UAV small object dataset VisDrone2019 and the remote sensing small object AITODv2, the network we designed achieved an accuracy index (mAP50) score of 48.9% and 49.1% with 1.3M parameters, successfully achieving a good balance between lightweight network and detection accuracy, achieving the best performance of lightweight models, and effectively reducing the occurrence of missed detections and false detections. The code will be available online at <span><span>https://github.com/Dzzz614/V2</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110471"},"PeriodicalIF":4.0,"publicationDate":"2025-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144335987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Technical study of innovative strategies of direct torque control for doubly fed induction motor — A review","authors":"Said Mahfoud ,&nbsp;Najib El Ouanjli ,&nbsp;Aziz Derouich","doi":"10.1016/j.compeleceng.2025.110498","DOIUrl":"10.1016/j.compeleceng.2025.110498","url":null,"abstract":"<div><div>Direct Torque Control (DTC) is widely recognized for its simplicity in both modeling and implementation. However, it still suffers from the major drawback of torque ripples. To address this limitation, numerous innovative solutions have been proposed in the literature. This paper provides an in-depth review of recent methods aimed at enhancing the performance of DTC, with a particular focus on optimizing Doubly-Fed Induction Motors (DFIM) in terms of torque and speed control. This review begins with a detailed presentation of the DFIM model, followed by a mathematical analysis of basic DTC control. The main challenges associated with this approach are identified and explained. The core section of the review explores a range of traditional techniques (such as backstepping and space vector modulation) as well as innovative methods based on optimization algorithms (genetic algorithm, ant colony optimization, and rooted tree optimization) and artificial intelligence, including fuzzy logic and neural networks, all aimed at improving DTC control efficiency. The objective is to reduce torque ripples while optimizing speed dynamics. Each of these techniques is analyzed in terms of its advantages and disadvantages, providing a critical perspective on their potential to enhance the performance of DTC control systems. This work stands out for its in-depth comparative study of these techniques based on major criteria (torque ripples and complexity), classification study and proposing actionable recommendations. This analysis aims to identify the most effective control strategies in the literature. The techniques that demonstrated the highest efficiency in this study are FL-DTC and ANN-DTC, which reduced torque ripples from 2.445 Nm for standard DTC to 1.14 Nm for FL-DTC and 1.08 Nm for ANN-DTC. Additionally, ANN-DTC offers the added benefit of lower complexity, providing a simpler yet equally effective solution compared to FL-DTC.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110498"},"PeriodicalIF":4.0,"publicationDate":"2025-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144329996","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A novel intrusion detection framework for industrial IoT: GCN-GRU architecture optimized with ant colony optimization","authors":"Mahdi Mir ,&nbsp;Mohammad Trik","doi":"10.1016/j.compeleceng.2025.110541","DOIUrl":"10.1016/j.compeleceng.2025.110541","url":null,"abstract":"<div><div>The swift proliferation of IIoT ecosystems has highlighted the essential requirement for effective Intrusion Detection System (IDS) to protect crucial infrastructures. This research presents a novel hybrid IDS that combines Graph Convolutional Networks (GCN) with Gated Recurrent Units (GRU), optimized by the Ant Colony Optimization (ACO) method, a bio-inspired meta-heuristic based on ant foraging behavior. This method automates hyperparameter adjustment, overcoming the constraints of conventional human optimization techniques. The proposed system utilizes GCN for structural feature extraction and GRU for sequential pattern analysis, facilitating thorough anomaly detection in IIoT traffic. The ACO-optimized IDS surpasses traditional optimization methods, including Genetic Algorithms, by attaining quicker convergence and enhanced performance metrics. Notwithstanding its effectiveness, the computational burden of the optimization approach necessitates additional enhancement. Experimental assessments of the EDGE-IIOTSET, CICAPT-IIoT, and WUSTL-IIoT datasets reveal detection accuracies of 97 % for the majority of attack scenarios, alongside improved scalability and diminished processing requirements. This study emphasizes the capability of integrating sophisticated neural architectures with nature-inspired optimization to enhance Industrial Internet of Things (IIoT) security against advancing cyber threats.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110541"},"PeriodicalIF":4.0,"publicationDate":"2025-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144329995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An optimized ECG copyright protection technique and its feature authentication","authors":"Ranjana Dwivedi ,&nbsp;Divyanshu Awasthi ,&nbsp;Vinay Kumar Srivastava","doi":"10.1016/j.compeleceng.2025.110546","DOIUrl":"10.1016/j.compeleceng.2025.110546","url":null,"abstract":"<div><div>The safe and dependable transfer of biomedical signals through smart medical devices is made possible by digital watermarking. This work proposes a reliable, secure, optimized watermarking technique for Electrocardiogram (ECG) data. Patient ID is used as watermark and pre-processed with 1-level lifting wavelet transform (LWT) before embedding. The suitable scaling factor is obtained using the Harris Hawks optimization (HHO) technique to balance the trade-off between robustness and imperceptibility. Watermark is encrypted using Henon map before embedding to provide security. QR decomposition and randomized singular value decomposition (RSVD) are applied to host ECG's horizontal and vertical sub-bands. Principal components (PC’s) are modified to embed the watermark instead of singular values, and thus, the proposed technique is free from false positive problems (FPP). The maximum Peak signal to noise ratio (PSNR) value obtained is 50.05 dB, and normalized correlation coefficient (NC) is 0.9975. The average percentage improvement in imperceptibility is 6.72 %, and in time complexity is 76.26 %. Comparison with existing ECG watermarking techniques establishes that the proposed ECG watermarking scheme outperforms in terms of robustness, imperceptibility, capacity, and computational complexity. Binary Robust Invariant Scalable Keypoints (BRISK) are used to verify the vital features of input ECG image to detect any undesired alteration.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110546"},"PeriodicalIF":4.0,"publicationDate":"2025-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144329997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deception and cloud integration: A multi-layered approach for DDoS detection, mitigation, and attack surface minimization in SD-IoT networks","authors":"Ameer El-Sayed ,&nbsp;Ahmed A. Toony ,&nbsp;Amr Tolba ,&nbsp;Fayez Alqahtani ,&nbsp;Yasser Alginahi ,&nbsp;Wael Said","doi":"10.1016/j.compeleceng.2025.110543","DOIUrl":"10.1016/j.compeleceng.2025.110543","url":null,"abstract":"<div><div>Detecting Distributed Denial of Service (DDoS) attacks in Software-Defined Internet of Things (SD-IoT) networks is challenging due to vulnerabilities in single-controller architectures, the limitations of the OpenFlow protocol, evolving DDoS strategies, and resource constraints. This research proposes a multi-layered security framework that integrates deception-based security, cloud-integrated machine learning (ML), a new hierarchically distributed multi-controller (HDMC) architecture, P4-enabled real-time traffic monitoring, and adaptive mitigation. The framework includes dynamic time-based windowing for enhanced detection, a decoy network to divert attackers, and a cloud-based multi-task ML model (MT-EDD) for attack classification. It also features a synchronized multi-control design for secure communication and coordinated actions among multiple controllers and a dynamic monitoring algorithm for real-time traffic analysis. P4 switches extract features from network traffic and send them to a cloud-based server for preprocessing and analysis by a pre-trained ensemble learning model (MT-EDD), which predicts attack states and communicates results to the central controller for mitigation. The controller then enforces appropriate mitigation actions on P4 switches. This approach offloads computationally intensive tasks to the cloud, improving scalability and detection accuracy. Evaluations show the framework achieves an average accuracy of 98.42%, precision of 96.17%, recall of 94.72%, F1-score of 95.39%, and specificity of 98.22%. The proposed P4-enabled solution consumes 30% less bandwidth and 25% less CPU, reduces detection times by 54.3%, and improves detection accuracy by 5.2% compared to the OpenFlow-enabled method. The HDMC architecture, evaluated against a single-controller setup, demonstrated 40% higher throughput and 32% lower latency, confirming its superior performance across multiple metrics.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110543"},"PeriodicalIF":4.0,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144321265","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integration of the layer 2 protection extended application for Open RAN: A security by design approach","authors":"Jihye Kim,&nbsp;Jaehyoung Park,&nbsp;Jong-Hyouk Lee","doi":"10.1016/j.compeleceng.2025.110479","DOIUrl":"10.1016/j.compeleceng.2025.110479","url":null,"abstract":"<div><div>The Open Radio Access Network (Open RAN) paradigm is progressively evolving to deliver open, intelligent, and innovative solutions for next-generation networks, particularly in 5G-Advanced and 6G. Spearheading this effort, the O-RAN Alliance has established a comprehensive framework for Open RAN and continues to lead standardization efforts in this domain. However, an analysis of the O-RAN security standard specifications reveals a primary focus on layer 3 and above, with an emphasis on IPsec protocols, leaving layer 2 security inadequately addressed. This gap exposes the E2 interface – a critical component enabling the openness and intelligence of Open RAN – to potential Man-in-the-Middle attacks at the layer 2. To mitigate this vulnerability, we propose a novel E2 protection xApp designed to enhance the security of the E2 interface. The proposed xApp detects and prevents layer 2 attacks, such as Address Resolution Protocol spoofing, ensuring robust protection for Open RAN environments. Experimental evaluations demonstrate that the proposed solution significantly improves network resilience, outperforming existing methods in mitigating layer 2 security vulnerability.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110479"},"PeriodicalIF":4.0,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144321267","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Physics-informed spatio-temporal network with trainable adaptive feature selection for short-term wind speed prediction","authors":"Laeeq Aslam ,&nbsp;Runmin Zou ,&nbsp;Yaohui Huang ,&nbsp;Ebrahim Shahzad Awan ,&nbsp;Sharjeel Abid Butt ,&nbsp;Qian Zhou","doi":"10.1016/j.compeleceng.2025.110517","DOIUrl":"10.1016/j.compeleceng.2025.110517","url":null,"abstract":"<div><div>Wind speed prediction (WSP) is essential for optimizing wind power generation, enhancing turbine performance and ensuring grid stability. Wind speed prediction models face challenges in effectively integrating complex spatio-temporal data with physical principles. This limitation reduces their accuracy and reliability for optimizing wind power generation and ensuring grid stability. To solve the issue, this study proposes a physics-informed spatio-temporal network (PISTNet) for short-term WSP that effectively integrates spatio-temporal data with physical principles. The proposed model designs a dynamic feature adapter (DFA) module that dynamically emphasizes relevant temporal and spatial information through adaptive masking mechanisms. It also incorporates an extended advection equation-based physical modeling (EPM) module, which provides physics-based WSP fused with neural network-extracted features using a feature fusion module (FFM). An adaptive physics penalty loss (APPL) function is proposed to enhance model’s accuracy to selectively enforce physical constraints based on significant prediction deviations. Comprehensive experiments conducted on four diverse datasets from Hamburg, Herning, Palmerston North, and Silkeborg demonstrate that the proposed model consistently outperforms six state-of-the-art prediction methods across multiple evaluation metrics, achieving up to a 7.1% improvement in RMSE, 8.0% in MAE, 2.3% in <span><math><mrow><mn>1</mn><mo>/</mo><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></mrow></math></span> score, and 9. 5% in SMAPE compared to the closest competitors. The findings highlight the potential of combining data-driven and physics-informed approaches to achieve accurate and reliable WSP, thereby contributing to the advancement of wind energy systems and grid management.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110517"},"PeriodicalIF":4.0,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144321266","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Battery capacity prediction for electric vehicles using an ensemble model in a fog computing framework","authors":"Yara A. Sultan ,&nbsp;Asmaa H. Rabie ,&nbsp;Amal Moharam ,&nbsp;Abdelfattah A. Eladl","doi":"10.1016/j.compeleceng.2025.110535","DOIUrl":"10.1016/j.compeleceng.2025.110535","url":null,"abstract":"<div><div>Electric Vehicles (EVs) are crucial in addressing environmental issues associated with traditional vehicles. However, accurately predicting battery capacity remains challenging for optimizing EV performance. This paper presents a Capacity Level Prediction Strategy (CLPS) consisting of three layers: Internet of Things (IoT), fog, and cloud. IoT data is initially sent to the fog layer for quick analysis and decision-making, then transferred to the cloud for long-term storage and future analysis. A Capacity Level Prediction Model (CLPM) is implemented in the fog layer, featuring two phases: preprocessing and capacity prediction. The preprocessing phase addresses missing data, outlier rejection, and feature selection. The prediction phase utilizes an Ensemble Prediction Model (EPM), combining the results of Random Forest (RF) and Deep Neural Network (DNN) models, with Logistic Regression (LR) for aggregation. The proposed CLPM outperforms existing approaches, achieving high accuracy (Mean Squared Error (MSE): 0.0003, coefficient of determination (R²): 0.9925, Mean Absolute Percentage Error (MAPE): 0.0066, and Mean Absolute Error (MAE): 0.01077639), and significantly improving battery monitoring, charging efficiency, and the overall lifespan of EVs.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110535"},"PeriodicalIF":4.0,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144314428","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Load balancing in cloud computing with multi-objective survivors optimization","authors":"R. Krishna Nayak ,&nbsp;G. Srinivasa Rao","doi":"10.1016/j.compeleceng.2025.110502","DOIUrl":"10.1016/j.compeleceng.2025.110502","url":null,"abstract":"<div><div>Cloud computing platform enables online services for data sharing, storage, and resource utilization to the cloud users. However, the major problem that occurs during cloud access is the server gets underloaded or overloaded affecting the processing time and resulting in the reduced quality of service (QoS). Specifically, the user tasks are allocated among the Virtual Machines (VMs) with diverse lengths, starting times, and processing times. Hence, load balancing is essential for ensuring that all the VMs are utilized appropriately. Consequently, this research proposes multi-objective optimization for load balancing while considering the network parameters such as makespan reduction, balanced CPU utilization, energy consumption minimization and throughput maximization. Specifically, the proposed MO-survivors’ optimization algorithm exploits the multi-objective fitness function considering the QoS constraints for selecting the VMs based on the capacity for achieving the parallel load execution. Further, the proposed algorithm effectively handles the network traffic, offers proper utilization of resources, manages the load capacity, and reduces the overprovision of infrastructure. The experimental outcomes reveals that the proposed MO-survivors’ optimization for load balancing exhibited better performance with 30 VMs attaining an improvement of 1.77 % over TSMGWO in terms of throughput, and attaining the makespan reduction of 223.03 s with TSMGWO. Further, the proposed approach revealed a reduced degree imbalance of 0.012 over TSMGWO and improved the resource utilization by 5.36 % compared to TSMGWO. Moreover, the results reveal the outstanding performance of the proposed MO-survivors optimization over the other existing algorithms used in the analysis.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110502"},"PeriodicalIF":4.0,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144314427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DiaXplain: A transparent and interpretable artificial intelligence approach for Type-2 diabetes diagnosis through deep learning","authors":"Sharandeep Singh ,&nbsp;Niyaz Ahmad Wani ,&nbsp;Ravinder Kumar ,&nbsp;Jatin Bedi","doi":"10.1016/j.compeleceng.2025.110470","DOIUrl":"10.1016/j.compeleceng.2025.110470","url":null,"abstract":"<div><div>Artificial intelligence has become a pivotal element in the healthcare industry, and it is used in healthcare administration, predictive modeling, decision-making, and diagnostics. Artificial intelligence technologies have achieved performance levels akin to humans in several activities; nevertheless, their widespread adoption is impeded by preconceptions of these systems as inscrutable “black boxes”. In response to such requirements, the authors devised <em>“DiaXplain”</em>, an interpretable hybrid artificial intelligence model intended to assist in diagnosing diabetes mellitus. <em>DiaXplain</em> utilizes data from the National Health and Nutrition Examination Survey (NHANES) and integrates a convolutional neural network (CNN) for feature extraction with XGBoost for classification, guaranteeing both elevated accuracy and interpretability. This hybrid methodology allows the model to autonomously discern pertinent properties while the XGBoost element enhances its prediction efficacy. <em>DiaXplain</em> utilizes SHAP (SHapley Additive exPlanations), an explainable artificial intelligence methodology that elucidates each prediction, assisting users in comprehending the reasoning behind individual and aggregate model actions. The model’s performance was assessed using necessary measures, revealing that <em>DiaXplain</em> exceeds current approaches in accuracy, attaining an exceptional 98.24%, with a precision of 95.12% and an F1-score of 97.50. Every prediction is substantiated by SHAP-generated explanations, providing doctors with a transparent understanding of the causes influencing each diagnostic result. Unlike traditional black-box systems, DiaXplain offers local and global interpretability, making its decisions understandable and trustworthy. This dual focus on high diagnostic accuracy and explainability makes DiaXplain a novel contribution to AI-driven healthcare, bridging the critical gap between prediction power and clinical transparency.This clarity is vital for cultivating confidence among healthcare providers, facilitating more informed decision-making, and enhancing the management of diabetes mellitus. <em>DiaXplain</em> signifies a significant progression in artificial intelligence-driven healthcare by merging high diagnostic accuracy with interpretable forecasts, providing healthcare practitioners with a dependable tool for diabetes diagnosis that promotes precision and fosters trust via transparency.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"126 ","pages":"Article 110470"},"PeriodicalIF":4.0,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144314429","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}