Chinese Journal of Electronics最新文献_第2页

A Novel GNN-Based Botnet Traffic Detection Approach Enhanced by LLM-Driven Data Augmentation 一种基于gnn的基于llm驱动数据增强的僵尸网络流量检测方法

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.405

Yan Hu;Yunhao Wang;Jiahao Huo;Xinyi Xu;Jun Zhu;Qiang Liu;Jian Wang

{"title":"A Novel GNN-Based Botnet Traffic Detection Approach Enhanced by LLM-Driven Data Augmentation","authors":"Yan Hu;Yunhao Wang;Jiahao Huo;Xinyi Xu;Jun Zhu;Qiang Liu;Jian Wang","doi":"10.23919/cje.2025.00.405","DOIUrl":"https://doi.org/10.23919/cje.2025.00.405","url":null,"abstract":"With the increasing interconnectivity among humans, machines, and networked devices, new architectures for production, communication, and service provision have emerged. However, many devices, notably legacy systems, contain numerous security vulnerabilities, which may be maliciously exploited during interconnection, potentially resulting in severe security incidents. Among major security threats, botnet attacks are particularly concerning. Traditional signature-based and behavior-based anomaly detection methods often overlook the topological dependencies among devices in complex communication networks, thereby creating a fundamental limitation for coordinated botnet-attack detection. Additionally, class imbalance in real-world datasets further limits detection performance. To address these challenges, we develop a multi-stage large language model (LLM)-driven data augmentation framework to synthesize high-fidelity minority class samples by learning intrinsic patterns from existing data. We extract fine-grained node and traffic features from the LLM-augmented traffic dataset. A graph neural network (GNN) then propagates and aggregates node features across the network to generate topology-aware node embeddings, which subsequently enhance the traffic features. Finally, a multilayer perceptron classifies the enhanced traffic features as normal or abnormal. We validate our approach through comprehensive experiments on the publicly available CTU-13 and UNSW-NB15 datasets, released by Czech Technical University and the University of New South Wales, Canberra, respectively. A comparative evaluation against multiple detection methods across diverse performance metrics demonstrates the superiority of our approach in botnet traffic detection.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"1876-1888"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188727","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Prompt Bins Learning for Monocular Depth Estimation 单目深度估计的提示箱学习

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.496

Yiru Wang;Zhibin Wang;Junxia Li;Qingshan Liu

{"title":"Prompt Bins Learning for Monocular Depth Estimation","authors":"Yiru Wang;Zhibin Wang;Junxia Li;Qingshan Liu","doi":"10.23919/cje.2025.00.496","DOIUrl":"https://doi.org/10.23919/cje.2025.00.496","url":null,"abstract":"Monocular depth estimation methods based on classification-regression have gained attention for their strong performance. These approaches predict depth via a linear combination of probability distributions and bin centers, where the design of depth bins critically influences accuracy. However, existing bin generation mechanisms often lack adaptability, exhibit low reliability, and incur high computational costs. To overcome these limitations, this paper introduces a prompt bins learning framework that dynamically generates depth bins using a learnable embedding mechanism. The proposed model enhances depth estimation by incorporating geometric priors through normal estimation, leveraging the intrinsic relationship between surface normals and depth to better capture scene structure. Specifically, prompt bins learn scene-specific depth distributions and reorganize them with decoded features to strengthen feature-prompt alignment. The proposed framework supports end-to-end training, enabling efficient optimization without the computational burden of traditional iterative methods. Experiments results on the standard indoor KITTI dataset and the outdoor NYU Depth v2 (NYU-v2) dataset demonstrate that the proposed approach achieves state-of-the-art performance across multiple metrics, confirming its effectiveness and robustness.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"2209-2222"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188655","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Deep Learning-Based Beam Training Methods for 6G MIMO Systems: From Far-Field to Near-Field 基于深度学习的6G MIMO系统波束训练方法：从远场到近场

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.308

Jikun Zhu;Bin Yan;Zheng Wang;Yongming Huang

{"title":"Deep Learning-Based Beam Training Methods for 6G MIMO Systems: From Far-Field to Near-Field","authors":"Jikun Zhu;Bin Yan;Zheng Wang;Yongming Huang","doi":"10.23919/cje.2025.00.308","DOIUrl":"https://doi.org/10.23919/cje.2025.00.308","url":null,"abstract":"Efficient beam training across the transition from conventional far-field massive multiple-input multiple-output (Massive MIMO) to near-field extremely large-scale MIMO (XL-MIMO) is fundamental to forth-coming sixth-generation (6G) systems, yet remains impeded by severe mobility, spherical-wave propagation and prohibitive training overhead. This paper presents a unified survey of contemporary deep learning (DL)-based beam training techniques, reviewing their evolution from early deep neural networks to more advanced architectures such as convolutional, recurrent, and other specialized models. These approaches exhibit strong adaptability and feature extraction capabilities. However, research in this area is still in its early stages, and such methods have yet to be widely applied in emerging scenarios, such as dynamically varying or multi-user interference-limited near-field communication systems. Motivated by the survey insights, we further introduce a lightweight stacked bidirectional gated recurrent unit (BiGRU) architecture equipped with residual shortcuts and adaptive normalization to capture the temporal evolution of near-field channels, referred to as the stacked residual-BiGRU (SRB). Simulations conducted on a dynamic XL-MIMO testbed demonstrate the competitive achievable rates and low training overhead of SRB, highlighting how the recurrent structure complements spatial feature extractors. Finally, we outline open challenges and future directions toward scalable, energy-efficient and robust beam management frameworks for real-world deployments.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"1889-1903"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A Parallel Sounding–Communication Integrated Method for MIMO Channels 一种MIMO信道并行测深通信集成方法

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.612

Dan Fei;He Hu;Haoran Chen;Runyu Han;Weirong Liu;Peng Zheng;Guoyu Wang;Chen Chen;Jiayi Zhang;Zhaohui Song;Bo Ai

{"title":"A Parallel Sounding–Communication Integrated Method for MIMO Channels","authors":"Dan Fei;He Hu;Haoran Chen;Runyu Han;Weirong Liu;Peng Zheng;Guoyu Wang;Chen Chen;Jiayi Zhang;Zhaohui Song;Bo Ai","doi":"10.23919/cje.2025.00.612","DOIUrl":"https://doi.org/10.23919/cje.2025.00.612","url":null,"abstract":"The deep synergy between high-precision channel measurement and reliable communication is pivotal for enhancing the performance of multiple-input multiple-output (MIMO) system. However, conventional approaches relying on orthogonal resource partitioning face significant bottlenecks in spectral efficiency and realtime capability, particularly due to the dedicated resource allocation required for pilots in large-scale antenna systems. To address these challenges, this paper proposes a parallel sounding-communication integrated method. The architecture achieves continuous channel sounding through three key innovations: a full-symbol reference signaling scheme that embeds reference sequences throughout the symbol duration to maximize sampling density; a scalable comb-type MIMO structure that ensures linear scalability and mitigates inter-antenna interference; and a flexible numerology mechanism capable of adapting to dynamic channel statistics. Simulation experiments conducted with standardized tapped delay line models demonstrate that the proposed method achieves a root mean square error vector magnitude of approximately 1.66% and power delay profile estimation with a maximum delay error below 0.16 μs. This approach effectively resolves the trade-off between sounding accuracy and signaling overhead, offering a scalable solution for future MIMO networks.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"2066-2073"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Spatial-Temporal Graph Neural Network Enabled Security Situation Prediction for 6G Edge Networks 基于时空图神经网络的6G边缘网络安全态势预测

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.489

Hanrui Xu;Wei Wang;Xiang Shao;Bo Zhou;Yiliang Liu;Xiaozhen Lu

引用次数: 0

Robust Set-Theoretic Action Quality Assessment in Complex Environments with Keypoint Uncertainty 具有关键点不确定性的复杂环境下鲁棒集论动作质量评估

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.317

Yun Zhu;Hao Wang;Shuang Liang;Xiaojun Wu;Maoguo Gong

{"title":"Robust Set-Theoretic Action Quality Assessment in Complex Environments with Keypoint Uncertainty","authors":"Yun Zhu;Hao Wang;Shuang Liang;Xiaojun Wu;Maoguo Gong","doi":"10.23919/cje.2025.00.317","DOIUrl":"https://doi.org/10.23919/cje.2025.00.317","url":null,"abstract":"Quantitative assessment of human motion quality has become critical across diverse domains including sports performance analysis, clinical rehabilitation monitoring, and human-robot interaction. However, automated evaluation faces significant challenges due to keypoint uncertainties caused by occlusion, motion blur, and detection failures. This paper introduces a novel action quality assessment method named dynamic action set comparator using evolutionary distance (DynASCEND) based on random finite set (RFS) theory. We model motion sequences as temporal realizations of evolving RFSs, explicitly addressing randomness in both observable keypoint cardinality and spatial states. An evolutionary distance metric is proposed to dynamically compare test and reference motions while robustly handling missing, extra, or ambiguous keypoints through optimal sub-pattern assignment. Extensive experiments on two specialized Chinese classical dance datasets (PANGTI and DIANFAN) demonstrate that DynASCEND exhibits strong robustness against keypoint perturbations, achieves high Pearson correlation with expert ratings, and performs comparably to state-of-the-art deep learning methods. Moreover, it offers distinct advantages including superior interpretability enabled by geometrically meaningful metrics and high computational efficiency, highlighting its practical utility in real-world applications that demand reliable and transparent motion quality assessment.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"2223-2241"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Channel Charting-Assisted Pilot Reuse for Uplink Extremely Large-Scale Multiple-Input Multiple-Output Transmission 链路超大规模多输入多输出传输的信道图辅助导频复用

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.303

Haohong Che;Li You;Jue Wang;Zhenzhou Jin;Chengjie Xie;Xiqi Gao

{"title":"Channel Charting-Assisted Pilot Reuse for Uplink Extremely Large-Scale Multiple-Input Multiple-Output Transmission","authors":"Haohong Che;Li You;Jue Wang;Zhenzhou Jin;Chengjie Xie;Xiqi Gao","doi":"10.23919/cje.2025.00.303","DOIUrl":"https://doi.org/10.23919/cje.2025.00.303","url":null,"abstract":"Extremely large-scale multiple-input multiple-output (XL-MIMO) is critical to sixth-generation communication networks. The substantial increase in the number of base station antennas gives rise to near-field propagation effects in wireless channels, which complicates channel parameter estimation and increases pilot overhead. Channel charting (CC) has emerged as a potent artificial intelligence (AI)-based technique to effectively extract features from varying high-dimensional channel statistics to enable pilot reuse (PR). In this paper, we investigate near-field channel estimation with reduced pilot overhead by developing a CC-assisted PR. To this end, we introduce a polar-domain codebook to capture the power distribution of near-field XL-MIMO channels. Based on these features, an intelligent CC network is developed to enable an effective low-dimensional mapping of the inherent correlation patterns in near-field user terminal (UT) channels. Building upon the mapped channel correlations, we further propose a near-field CC-assisted pilot allocation (NCC-PA) algorithm, which efficiently enhances channel orthogonality among pilot-reusing UTs. Numerical results confirm that the NCC-PA algorithm substantially elevates the wireless transmission performance, offering a marked improvement over the conventional non-AI approach and far-field PR scheme.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"1952-1959"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Marine Traffic Flow Prediction Based on Kolmogorov-Arnold Network Using Multiscale Features 基于多尺度特征Kolmogorov-Arnold网络的海上交通流预测

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.414

Feifei Zhou;Xiangling Li;Xu Chen;Fengyuan Zhang

{"title":"Marine Traffic Flow Prediction Based on Kolmogorov-Arnold Network Using Multiscale Features","authors":"Feifei Zhou;Xiangling Li;Xu Chen;Fengyuan Zhang","doi":"10.23919/cje.2025.00.414","DOIUrl":"https://doi.org/10.23919/cje.2025.00.414","url":null,"abstract":"Accurate prediction of marine traffic flow is crucial for ensuring maritime safety. However, marine traffic flow exhibits nonlinearity and non-stationarity, posing significant challenges to the performance improvement of deep learning based prediction models. To address these challenges, we propose a novel composite model by combining Kolmogorov-Arnold network (KAN) with variational mode decomposition (VMD) and bidirectional long short-term memory network (BiLSTM), termed VMD-multi-feature BiLSTM (MFBiLSTM)-KAN model. The original non-stationary time series are decomposed by VMD. We generate multiscale features with multiscale information by concatenating original time series with stable time series generated by VMD. Then, BiLSTM captures bidirectional dependency patterns among multiscale features. KAN flexibly maps the outputs of BiLSTM to final prediction results, by using learnable activation functions as weights. We employ the VMD-MFBiLSTM-KAN model for marine traffic flow prediction and validate its performance using real-world datasets derived from automatic identification system. Experimental results demonstrate that the VMD-MFBiLSTM-KAN model achieves superior prediction accuracy compared with other traditional models by utilizing multiscale features with comprehensive information and KAN with high flexibility.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"2085-2098"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188761","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

OAAH: Offloaded Anonymous Authentication and Handover Protocol for Satellite-Integrated Vehicular Networks 卫星集成车辆网络的卸载匿名认证和切换协议

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.351

Feifei Liu;Jianwei Liu

{"title":"OAAH: Offloaded Anonymous Authentication and Handover Protocol for Satellite-Integrated Vehicular Networks","authors":"Feifei Liu;Jianwei Liu","doi":"10.23919/cje.2025.00.351","DOIUrl":"https://doi.org/10.23919/cje.2025.00.351","url":null,"abstract":"Satellite-integrated vehicular networks (SVNs) extend the coverage of terrestrial networks to vehicles in remote or extreme environments. In such highly dynamic and mobile settings, secure authentication and seamless handover are essential for reliable services. However, SVNs inherently suffer from high propagation delays, frequent handovers, and severe exposure to security and privacy threats. Among these risks, fake satellite attacks represent a critical yet largely overlooked risk, and existing schemes fail to address these challenges in a unified manner. To this end, we propose OAAH, an offloaded anonymous authentication and handover protocol for SVNs, which enables vehicles to authenticate anonymously to ground stations (GSs) via low Earth orbit (LEO) satellites. OAAH offers three key advantages: attribute-based anonymous authentication and handover with selective disclosure; efficient handover that reduces vehicle-side computation by 93.2% compared to initial authentication when undisclosing 30 attributes; and it is the first protocol against fake satellite attacks while ensuring strong security and privacy guarantees, including attribute privacy, vehicle anonymity, unlinkability, and accountability. Extensive experimental results demonstrate that OAAH achieves practical performance with minimal overhead, making it a promising solution for privacy-preserving authentication in SVNs.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"2158-2171"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An Area-Efficient Power Clamp Circuit for Surge Protections 一种用于浪涌保护的面积高效电源钳位电路

IF 4.5 4区计算机科学

Chinese Journal of Electronics Pub Date : 2026-09-01 Epub Date: 2026-10-16 DOI: 10.23919/cje.2025.00.369

Ran Ye;Zhiqiang Hu;Ke Cui;Qiao Kang;Hao Luo;Weipeng Ye;Siyang Liu;Weifeng Sun

{"title":"An Area-Efficient Power Clamp Circuit for Surge Protections","authors":"Ran Ye;Zhiqiang Hu;Ke Cui;Qiao Kang;Hao Luo;Weipeng Ye;Siyang Liu;Weifeng Sun","doi":"10.23919/cje.2025.00.369","DOIUrl":"https://doi.org/10.23919/cje.2025.00.369","url":null,"abstract":"Surge, characterized by higher energy and longer duration, has a serious threat to semiconductor devices, components and systems. The metal oxide semiconductor field effect transistor (MOSFET)-based power clamp circuits are widely used for surge protection devices due to their ability to suppress overvoltage and discharge current during surge events. In this work, an areaefficient power clamp circuit is proposed and fabricated based on the 0.18 μm bipolar-CMOS-DMOS process. The conventional diode string is replaced to optimize the chip area. The experimental results show that the proposed circuit stably clamps the surge voltage at a certain value and provides reliable surge protection. Compared to the conventional diode-triggered circuits, the proposed power clamp circuits achieve an area reduction of 9.1% and 14.2% for clamping voltages of 33 V and 54 V, respectively. The peak surge current, leakage current and output capacitance are measured and show no significant deviation from those of conventional circuits. Furthermore, the results of reliability experiments demonstrate that the circuit can maintain operational stability and performance integrity under the conditions of multiple surge impact and high temperature.","PeriodicalId":50701,"journal":{"name":"Chinese Journal of Electronics","volume":"35 5","pages":"1960-1969"},"PeriodicalIF":4.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149188264","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0