检索结果-内蒙古大学图书馆

Robust federated contrastive recommender system against targeted model poisoning attack

Science China(Information Sciences) 2025年第4期68卷 50-65页

作者： Wei YUAN Chaoqun YANG Liang QU Guanhua YE Quoc Viet Hung NGUYEN Hongzhi YIN School of Electrical Engineering and Computer Science The University of Queensland School of Information and Communication Technology Griffith University Deep Neural Computing Company Limited

Federated recommender systems(FedRecs) have garnered increasing attention recently, thanks to their privacypreserving benefits. However, the decentralized and open characteristics of current FedRecs present at least two ***, the performance of FedRecs is compromised due to highly sparse on-device data for each client. Second, the system's robustness is undermined by the vulnerability to model poisoning attacks launched by malicious users. In this paper, we introduce a novel contrastive learning framework designed to fully leverage the client's sparse data through embedding augmentation, referred to as CL4FedRec. Unlike previous contrastive learning approaches in FedRecs that necessitate clients to share their private parameters, our CL4FedRec aligns with the basic FedRec learning protocol, ensuring compatibility with most existing FedRec implementations. We then evaluate the robustness of FedRecs equipped with CL4FedRec by subjecting it to several state-of-the-art model poisoning attacks. Surprisingly, our observations reveal that contrastive learning tends to exacerbate the vulnerability of FedRecs to these attacks. This is attributed to the enhanced embedding uniformity, making the polluted target item embedding easily proximate to popular items. Based on this insight, we propose an enhanced and robust version of CL4FedRec(rCL4FedRec) by introducing a regularizer to maintain the distance among item embeddings with different popularity levels. Extensive experiments conducted on four commonly used recommendation datasets demonstrate that rCL4FedRec significantly enhances both the model's performance and the robustness of FedRecs.

关键词： federated recommender system contrastive learning model poisoning attack and defense

来源：评论

学校读者我要写书评

暂无评论

Photonic generation of multi-format and reconfigurable microwave signals based on cascaded external modulation

引用

Optical and Quantum Electronics 2025年第3期57卷 1-14页

作者： Chen, Mo Chen, Hongyao Zhang, HongYu Li, ZhuoYang Wang, Jianping Feng, LiFang Lu, HuiMin School of Computer and Communication Engineering University of Science and Technology Beijing Beijing10083 China

This paper presents a photonic scheme for generating multi-format, multi-band, and reconfigurable microwave photonic signals through cascaded external modulation. The proposed system utilize dual-parallel Mach–Zehnder modulators (DP-MZM) and Mach–Zehnder modulators (MZM) to produce high-order optical sidebands, enabling flexible frequency multiplication factors switching and pulse waveform reconstruction. Additionally, using a polarization control structure allows for spurious suppression and encoded signal loading. The system can generate various signal formats, including rectangular optical frequency combs, frequency-multiplied signals, dual-frequency encoded signals, and multiple pulse waveforms. Simulation results demonstrate the successful generation of 16 GHz frequency-doubled signal and 32 GHz frequency-quadrupled signal, with spurious suppression ratios of 37.9 dB and 44.8 dB, respectively. It can also generate four-line rectangular optical frequency combs with frequency multiplication factors of 2 and 4, achieving flatness values of 0.24 dB and 0.16 dB. Furthermore, the system can generate symmetric triangular pulses and Sinc-Nyquist pulses with adjustable repetition frequencies, as well as encoded signals with good pulse compression performance, having PCR values close to the theoretical value of 13. This work demonstrates a versatile and reconfigurable approach to microwave photonic signal generation, offering promising applications in advanced radar systems. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2025.

关键词： Microwave photonics

来源：评论

学校读者我要写书评

暂无评论

A ranking framework for the selection of IoT cloud platforms using hybrid multi-attribute decision-making method

引用

International Journal of Intelligent Computing and Cybernetics 2024年第4期17卷 824-843页

作者： Supriya Raheja Rakesh Garg Ritvik Garg Department of Computer Science and Engineering Amity University Uttar PradeshNoidaIndia Department of Computer Science and Engineering Gurugram UniversityGurugramIndia Department of Computer and Communication Engineering Manipal University JaipurJaipurIndia

Purpose-The Internet of Things(IoT)cloud platforms provide end-to-end solutions that integrate various capabilities such as application development,device and connectivity management,data storage,data analysis and data *** high use of these platforms results in their huge availability provided by different ***,choosing the optimal IoT cloud platform to develop IoT applications successfully has become *** key purpose of the present study is to implement a hybrid multi-attribute decision-making approach(MADM)to evaluate and select IoT cloud ***/methodology/approach-The optimal selection of the IoT cloud platforms seems to be dependent on multiple ***,the optimal selection of IoT cloud platforms problem is modeled as a MADM problem,and a hybrid approach named neutrosophic fuzzy set-Euclidean taxicab distance-based approach(NFS-ETDBA)is implemented to solve the ***-ETDBA works on the calculation of assessment score for each alternative,*** cloud platforms,by combining two different measures:Euclidean and taxicab ***-A case study to illustrate the working of the proposed NFS-ETDBA for optimal selection of IoT cloud platforms is *** results obtained on the basis of calculated assessment scores depict that“Azure IoT suite”is the most preferable IoT cloud platform,whereas“Salesman IoT cloud”is the least ***/value-The proposed NFS-ETDBA methodology for the IoT cloud platform selection is implemented for the first time in this *** is highly capable of handling the large number of alternatives and the selection attributes involved in any decision-making ***,the use of fuzzy set theory(FST)makes it very easy to handle the impreciseness that may occur during the data collection through a questionnaire from a group of experts.

关键词： Internet of things(IoT) Multi-attribute decision-making Neutrosophic fuzzy set Euclidean and taxicab distance

来源：评论

学校读者我要写书评

暂无评论

Learning-based user association and dynamic resource allocation in multi-connectivity enabled unmanned aerial vehicle networks

引用

Digital communications and Networks 2024年第1期10卷 53-62页

作者： Zhipeng Cheng Minghui Liwang Ning Chen Lianfen Huang Nadra Guizani Xiaojiang Du Department of Information and Communication Engineering Xiamen UniversityXiamen361005China School of Electrical and Computer Engineering University of Texas at ArlingtonArlingtonTX76019USA Department of Electrical and Computer Engineering Stevens Institute of TechnologyHobokenNJ07030USA

Unmanned Aerial Vehicles(UAvs)as aerial base stations to provide communication services for ground users is a flexible and cost-effective paradigm in ***,dynamic resource allocation and multi-connectivity can be adopted to further harness the potentials of UAVs in improving communication capacity,in such situations such that the interference among users becomes a pivotal disincentive requiring effective *** this end,we investigate the Joint UAV-User Association,Channel Allocation,and transmission Power Control(J-UACAPC)problem in a multi-connectivity-enabled UAV network with constrained backhaul links,where each UAV can determine the reusable channels and transmission power to serve the selected ground *** goal was to mitigate co-channel interference while maximizing long-term system *** problem was modeled as a cooperative stochastic game with hybrid discrete-continuous action space.A Multi-Agent Hybrid Deep Reinforcement Learning(MAHDRL)algorithm was proposed to address this *** simulation results demonstrated the effectiveness of the proposed algorithm and showed that it has a higher system utility than the baseline methods.

关键词： UAV-user association Multi-connectivity Resource allocation Power control Multi-agent deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

SCGformer: Smart contract vulnerability detection based on control flow graph and transformer

IET Blockchain

引用

IET Blockchain 2023年第4期3卷 213-221页

作者： Gong, KeXin Song, Xiangmei Wang, Na Wang, Chunyang Zhu, Huijuan School of Computer Science and Communication Engineering Jiangsu University Zhenjiang China

The security of smart contract has always been one of the significant problems in blockchain. As shown in previous studies, vulnerabilities in smart contracts can lead to unpredictable losses. With the rapid growth of the number of smart contracts, more and more data driven detection technologies based on machine learning have been proposed. However, some state-of-the-art approaches mainly rely on the source code of smart contract. These methods are limited by the openness of the source code and the version of the programming language. To address this problem, we propose a novel vulnerability detection method based on transformer by constructing the control flow graph (CFG) of smart contracts operation codes (opcodes), which shields the difference of various versions of program language. Extensive experiments are conducted to evaluate the effectiveness of the proposed method on the authors' own collected dataset. The experimental results show that the proposed method achieves 94.36% accuracy in vulnerability detection, which performs better than other state-of-the-art methods. © 2023 The Authors. IET Blockchain published by John Wiley & Sons Ltd on behalf of The Institution of engineering and Technology.

关键词： Smart contract

来源：评论

学校读者我要写书评

暂无评论

Security to text (S2T): multi-layered based security approaches for secret text content

引用

Multimedia Tools and Applications 2025年第15期84卷 15151-15171页

作者： Kashid, Shamal Awasthi, Lalit K. Berwal, Krishan Department of Computer Science and Engineering National Institute of Technology Uttarakhand IN Srinagar India Faculty of Communication Engineering Military College of Telecommunication Engineering IN Mhow India

In the digital world, text data is produced in an unstructured manner across various communication channels. Extracting valuable information from such data with security is crucial and requires the development of techniques in text mining, information retrieval, and natural language processing (NLP). To solve this issue, we introduce two novel approaches: keyword extraction (KE) and a multi-layered secret sharing scheme (MLSS) to provide security to extracted keywords rather than overall text documents. The KE approach encompasses a sequence of text pre-processing procedures, including tokenization, stopword removal, stemming, and bag of words representation, followed by indexing. This methodology aims to efficiently extract keywords from text datasets. For this research work, we have proposed three datasets, including text messages, whatsapp messages, and electronic mail. MLSS enhances the security of extracted textual data by leveraging text pre-processing steps. This scheme ensures better confidentiality and the non-revealment of sensitive information. Additionally, we evaluate our KE model on our dataset as well as on standard datasets. Experimental results demonstrate the effectiveness of our proposed security to text (S2T) model, which outperform existing state-of-the-art approaches. The model obtains a 100% correlation between the reconstructed text and the original text. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

A two-stage CNN model for the classification and severity analysis of retinal and choroidal diseases in OCT images

引用

International Journal of Intelligent Networks 2024年第1期5卷 10-18页

作者： George, Neetha Shine, Linu N, Ambily Abraham, Bejoy Ramachandran, Sivakumar Department of Electronics and Communication Engineering Rajiv Gandhi Institute of Technology Kottayam Kerala India Department of Electronics and Communication Engineering College of Engineering Trivandrum Kerala India Department of Electronics and Communication Engineering Government Engineering College Idukki Kerala India Department of Computer Science and Engineering College of Engineering Muttathara Kerala Thiruvananthapuram India Department of Electronics and Communication Engineering Government Engineering College Wayanad Kerala India

The advancements in medical imaging techniques have brought exponential increase in the quantity and complexity of data which often require human expertise for interpretation and decision making. However, in real-world clinical settings, there is often a shortage of experts available for timely diagnosis and triage. An automated diagnostic technique aids clinicians in the precise diagnosis and effective management of diseases. In this article, a two-stage classification model using convolutional neural network (CNN) is proposed for the classification and severity analysis of retinal and choroidal diseases in optical coherence tomography (OCT) images. The proposed model can identify abnormal conditions such as Pachychoroid disorders, macular edema, and Drusen. The images are initially classified into four categories-healthy, Drusen, Pachychoroid and macular edema classes. The severity of Pachychoroid conditions, including central serous chorioretinopathy, polypoid choroidal vasculopathy, and choroidal neovascularization, is subsequently determined through a second-level classification. A modified version of the VGG16 architecture is used for the initial classification. A fine-tuned CNN with the same architecture is then employed to determine the severity of Pachychoroid diseases. Further, we make use of the UNet architecture for assessing the severity of macular edema. The proposed hybrid approach for classification and analysis achieved promising results, demonstrating consistency on par with human experts in diagnosing and grading ocular diseases. © 2024

关键词： Optical tomography

来源：评论

学校读者我要写书评

暂无评论

Preamble detection for broadband power line communication systems

引用

Telecommunication Systems 2024年第1期87卷 47-60页

作者： Su, Szu-Lin Huang, Nan-Hsiung Chen, Yu-Hsuan Lai, Kuei-Chiang Institute of Computer and Communication Engineering Department of Electrical Engineering National Cheng Kung University Tainan Taiwan

Power line communication (PLC) systems enable information exchange over existing electric power grids at a low deployment cost. Such systems constitute a crucial component of smart grids. However, PLC is negatively affected by multipath channel, random impulsive noise (IN), and background noise. Numerous studies have evaluated the data detection performance of PLC systems;however, few studies have examined the preamble detection performance of these systems under IN. Accordingly, the present study evaluated preamble detection schemes for PLC systems. Specifically, the study compared two schemes, namely, an energy detection scheme and an autocorrelation detection scheme. The study results indicated that the energy detection scheme exhibited lower complexity and had a shorter mean time to correct detection than did the autocorrelation detection scheme. Therefore, the energy detection scheme was determined to be more suitable than the autocorrelation detection scheme for preamble detection in power line channels, a finding that is in contrast to expectations. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Impulse noise

来源：评论

学校读者我要写书评

暂无评论

A review on Gujarati language based automatic speech recognition (ASR) systems

引用

International Journal of Speech Technology 2024年第1期27卷 133-156页

作者： Dua, Mohit Bhagat, Bhavesh Dua, Shelza Chakravarty, Nidhi Department of Computer Engineering National Institute of Technology Kurukshetra India Department of Electronics Communication and Engineering Punjab Engineering College Chandigarh Chandigarh India

Automatic speech recognition (ASR) plays a crucial role in facilitating natural and efficient human–computer interaction. This paper offers a comprehensive review of ASR systems tailored specifically for the Gujarati language. Existing literature on Gujarati ASR indicates that most survey papers have focused solely on back-end classification methods. This paper aims to fill this gap by presenting a comprehensive survey that encompasses feature extraction techniques, backend models, speech datasets, and evaluation metrics. This review provides an in-depth analysis of the cutting-edge methodologies employed in the development of Gujarati ASR systems. Firstly, the paper discusses various feature extraction techniques. Secondly, it covers classification models and their impact on ASR performance. Thirdly, the study delves into available speech datasets relevant to the Gujarati language, offering valuable insights for researchers and practitioners. Finally, the paper reviews the evaluation parameters used in ASR systems. Additionally, it gives an overview of online toolkits, resources, and language models pertinent to Gujarati ASR. This review serves as a comprehensive reference for academics and professionals involved in the research and development of ASR systems for the Gujarati language. The parameters selected for comparison include front-end feature extraction methods, back-end classification techniques, speech dataset, and evaluation metrics. Furthermore, the paper discusses both the contributions and limitations encountered by current ASR systems during the review. Finally, it addresses various challenges that still persist and provides directions for future research in this critical field. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

Joint Resource Allocation and Trajectory Optimization Algorithm in UAV-Assisted MEC System Based on DDPG

Joint Resource Allocation and Trajectory Optimization Algori...

引用

2024 International Annual Conference on Complex Systems and Intelligent Science, CSIS-IAC 2024

作者： Song, Xin Li, Ruomeng Xu, Siyang Zhou, Zijian School of Computer and Communication Engineering Northeastern University at Qinhuangdao Qinhuangdao China School of Computer and Communication Engineering Northeastern University Shenyang China

ISBN: (纸本)9798331504755

Due to the advantages of maneuverability and the small size of Unmanned Aerial Vehicles (UAVs), mobile edge computing (MEC) servers can be deployed for them to provide computing services. Because the computing resources and energy are limited and the tasks have strict time constraints, tasks are prioritized and their age of information (AoI) is recorded. To solve the issues, we propose a deep deterministic policy gradient (DDPG)-based joint resource allocation and trajectory optimization algorithm in UAV-assisted MEC system. We formulate the resource allocation and trajectory problems as a Markov Decision Process (MDP). Then, we propose a DDPG algorithm to solve the formulated problem and minimize the total energy consumption of the system. Simulation results show that the proposed DDPG-based joint optimization algorithm significantly reduces the energy consumption of the system and has a fast convergence rate. © 2024 IEEE.

关键词： Markov processes

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：