检索结果-内蒙古大学图书馆

Developed a Machine Learning and Deep Learning Model for 5G MIMO Data Based Beam Selection and Intelligent Network Analytics

引用

SN computer Science 2025年第3期6卷 1-21页

作者： Ashok, P. Suganya, R. Lekha, P. Pandian, Poornima Subashini, M. Nithya, M. Department of Computer Science and Engineering Sri Sai Ram Institute of Technology Tamilnadu India Department of Computer and Communication Engineering Sri Sai Ram Institute of Technology Tamilnadu India Department of Electronics and Communication Engineering Sri Sairam Engineering College Tamilnadu India Department of Computer Science and Engineering Sri Sairam Engineering College Tamilnadu India

Introduces an innovative application of 5G MIMO data in singular value decomposition (SVD) focusing on deep learning beam-selection and intelligent network analytics. This research explores the potential of leveraging 5G MIMO data to enhance network performance and efficiency through advanced ML techniques. Deep learning algorithms are employed for beam-selection, optimizing the transmission of data in 5G networks to improve throughput and reliability. Intelligent network analytics are applied to extract valuable insights from the vast amount of data generated by 5G MIMO systems, enabling proactive network management and optimization. By harnessing the capabilities of deep learning and intelligent analytics, this research aims to address key challenges in 5G network optimization and management. The integration of SVD techniques with 5G MIMO data offers a novel approach to enhancing network performance, enabling more efficient utilization of resources and improved quality of service for users. The insights gained from intelligent network analytics facilitate informed decision-making and proactive maintenance leading to greater reliability and resilience in 5G networks. This research contributes to the advancement of 5G technology by leveraging ML to unlock the full potential of MIMO data for optimizing network performance and enabling intelligent network management. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2025.

关键词： 5G Beam-selection Deep learning Intelligent management MIMO Network analytics Network performance Optimization Proactive maintenance Singular value decomposition

来源：评论

学校读者我要写书评

暂无评论

Two-dimensional confined topological modes in all dielectric guided mode resonance structure through surface features

引用

Optical and Quantum Electronics 2025年第4期57卷 1-17页

作者： Ramachandra, Swathy Menezes, Maclean Paul Tumuluru, Vamsi Krishna Kulkarni, Raghavendra G. Sinha, Rajat K. Bhowmick, Kaustav Department of Electronics and Communication Engineering PES University Karnataka Bangalore India Department of Electrical and Computer Engineering University of Toronto Toronto Canada

Guided mode resonance (GMR) structures offer simplicity and have spectral resonance capabilities such as narrowband resonance, high Q-factor, whereas topological surface features offer spatial light control. In this paper, we explore the integration of the GMR structures with topological photonics to achieve extreme two-dimensional (2D) spatial confinement of light. With this integration, we propose a novel design that leverages discontinuous surface features (crosscut) to excite Jackiw—Rebbi (JR) solution, which describes relationship between Dirac equation and topological insulators. Finite element modelling, considering various structural and material parameters, demonstrate a 40% improvement in light confinement compared to benchmark GMR structures, with a calculated full width half maximum (FWHM) of the confined optical mode measured to be approximately ~ 18 ± 3 nm. This FWHM value represent the spatial extent of the confined optical field, indicating the achieved field localization in the proposed structures. This significant enhancement showcases the potential of our approach for applications demanding ultra- precise spatial resolution. Furthermore, tolerance studies reveal good robustness to variations in the dimensions of the crosscut and angle of polarization, ensuring practical feasibility. The proposed structure exhibited good tolerance of up to ~ ± 10%. This work paves the way for highly confined 2D light manipulation within all- dielectric platforms, opening exciting avenues for nanophotonic devices and applications requiring light localization. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2025.

关键词： Topology

来源：评论

学校读者我要写书评

暂无评论

Intelligent Medical Literature Access for Healthcare Professionals Using NLP and Pubmed 2

Intelligent Medical Literature Access for Healthcare Profess...

引用

2nd IEEE International Conference on Emerging Research in Computational Science, ICERCS 2024

作者： Sangeetha, R. Karishma, S.M. Kiran Kumar, A. Lordson Gnanadurai, A. Prabanjan, R. Sri Eshwar College of Engineering Department of Computer and Communication Engineering Coimbatore India

ISBN: (纸本)9798331534950

MEDITRON represents a transformative approach in healthcare informatics by combining advanced natural language processing (NLP) capabilities with the extensive PubMed database. This integration enables real-time, context-sensitive responses to diverse user queries, serving the distinct needs of healthcare professionals, researchers, and patients. Unlike traditional keyword-based searches, which often yield overwhelming or irrelevant information, MEDITRON is designed to understand the context and specific requirements of each user type, offering precise, tailored information in real time. Healthcare professionals need quick access to relevant studies for patient care, researchers seek detailed publications for academic purposes, and patients require accessible information on their health conditions. MEDITRON's sophisticated NLP technology processes these varied queries accurately, translating complex medical literature into user-appropriate formats and its intuitive user interface and robust filtering features, MEDITRON is built to be user-friendly while providing high accuracy in search results. This tool minimizes time spent sifting through large volumes of information, reducing information overload and improving the efficiency of information retrieval. By empowering users with timely, relevant data, MEDITRON aims to redefine the accessibility of medical literature, promoting informed decision-making, enhancing knowledge dissemination, and ultimately improving healthcare outcomes. In essence, MEDITRON exemplifies how healthcare informatics can evolve to meet the needs of a diverse user base by aligning the power of NLP with the comprehensive reach of the PubMed database. © 2024 IEEE.

关键词： By empowering users with timely information reducing information overload and improving the efficiency of information retrieval relevant data

来源：评论

学校读者我要写书评

暂无评论

Multimodal Approach for Diabetic Retinopathy Detection Using Deep Learning and Clinical Data Fusion 9

Multimodal Approach for Diabetic Retinopathy Detection Using...

引用

9th International Conference on communication and Electronics Systems, ICCES 2024

作者： Karthikeyan, S. Selva Sangeeth, S. Sreeja, G.G. Sivasanjeev, R. Srimathi, M. Sri Eshwar College of Engineering Department of Computer and Communication Engineering Coimbatore India

ISBN: (数字)9798350377972

ISBN: (纸本)9798350377972

A novel multimodal diabetic retinopathy (DR) diagnosis method is proposed that addresses the most common and time-consuming limitations of traditional diagnosis methods. This approach combines deep learning techniques with medical data fusion to integrate retinal images with important measurements such as glucose levels, blood pressure, and diabetes duration. The model uses convolutional neural networks (CNN) for high-level image analysis and multilayer perceptrons (MLP) to process medical data, and achieves fusion through integration. Compared with image-only models, this principle improves performance metrics including optimization, recall, and classification accuracy. Through the use of visualization and therapy, this approach provides large-scale and flexible solutions for many clinical environments. This study demonstrates the various aspects of multidisciplinary work in the diagnosis of DR, paving the way for better patient outcomes and setting a new standard for integration into expert clinical practice in medicine. Future work will focus on expanding the data, developing the translational model and exploring real-world applications. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Quad Band Split Octagonal Ring Antenna with Integrated Stub for Satellite communication-Dependent Wireless Applications

引用

Progress In Electromagnetics Research C 2024年 141卷 195-204页

作者： Suganthi, Jambulingam Kavitha, Thamizhchelvan Department of Electronics & Communication Engineering PES University Karnataka Bangalore560085 India Department of Computer Engineering New Horizon College of Engineering Karnataka Bangalore560103 India

The purpose of this study is to design a multiband antenna using metamaterial for efficient satellite communication. The majority of the antennae described in the available research suffer from a variety of limitations, including intricate designs, great footprints, and erratic radiation patterns. Therefore, there is a significant demand for antennae that are of a smaller size but nevertheless perform well. This paper proposes a quad-band stub-incorporated split octagonal ring antenna for satellite communication-dependent wireless applications. The suggested antenna is built on an FR4 substrate that measures 22 × 39 × 1.6 mm3. CST EM studio software is used for the entire simulation. The proposed antenna resonates at four different bands, with operating frequencies ranging from 2.15 GHz to 2.30 GHz, 2.86 GHz to 3.76 GHz (due to stub 1), 4.47 GHz to 5.24 GHz (due to stub 2), and 5.67 GHz to 6.35 GHz (due to stub 3). (due to gap between the stub). The proposed antenna has resonant frequencies of 2.23 GHz, 3.28 GHz, 4.77 GHz, and 5.89 GHz, and bandwidths of 153 MHz, 9011 MHz, 7692 MHz, and 6813 MHz. Parametric analysis is used to select the best values. The designed antenna is built and tested. The measured and simulated values for return loss, gain, E-plane, and H-plane are compared, and they agree. Its dual-band operation, compact size, steady radiation pattern, and gain above 1 dBi across the whole resonating band make it suited for ISM, WIFI, WLAN, WIMAX, 5G, and C band satellite applications. © 2024, Electromagnetics Academy. All rights reserved.

关键词： Directional patterns (antenna)

来源：评论

学校读者我要写书评

暂无评论

Design and Implementation of Med Vault - Biometric-based Medical Record System 5

Design and Implementation of Med Vault - Biometric-based Med...

引用

5th International Conference on Electronics and Sustainable communication Systems, ICESC 2024

作者： Selva Sangeeth, S. Srinivasan, Dhamodharan Karthikeyan, S. Sivasanjeev, R. Srimathi, M. Sri Eshwar College of Engineering Department of Computer and Communication Engineering Coimbatore India

ISBN: (纸本)9798350379945

The 'Med Vault - Biometric Based Medical Record System' is an innovative solution proposed in this research study to streamline medical records management using biometric technology. Despite its promising concept, several aspects need improvement. Firstly, the reliance on biometric authentication raises privacy and security concerns. Although biometrics provide a convenient way to verify identity, they are not foolproof. Hackers could potentially gain access to sensitive medical information by bypassing or manipulating the biometric authentication process. Therefore, the developers of Med Vault must invest in robust encryption and security measures to protect patient data. Additionally, the system's effectiveness heavily relies on widespread adoption by healthcare providers. Without universal implementation, patients may still face challenges when accessing their medical records across different healthcare facilities. It would be beneficial for Med Vault to partner with major healthcare organizations and incentivize them to integrate this system into their existing infrastructure. Furthermore, while Med Vault claims to improve efficiency by reducing paperwork and administrative tasks, it fails to address potential issues related to data entry errors or incomplete records. The accuracy and completeness of medical records are essential for providing quality care, and any system aiming to revolutionize record-keeping should prioritize these aspects. In conclusion, while the 'Med Vault - Biometric Based Medical Record System' presents an exciting opportunity for improving medical record management, significant areas require attention before its widespread implementation can be considered successful. Addressing privacy concerns, ensuring universal adoption among healthcare providers, and prioritizing the accuracy and completeness of records will be key factors in determining its long-term viability as a solution for efficient medical record management. © 2024 IEEE.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

A transformer based real-time photo captioning framework for visually impaired people with visual attention

引用

Multimedia Tools and Applications 2024年第41期83卷 88859-88878页

作者： Muhammed Kunju, Abubeker Kiliyanal Baskar, S. Zafar, Sherin A R, Bushara S, Rinesh A, Shafeena Karim Kerala Kanjirapally India Faculty of Engineering Department of Electronics and Communication Engineering Karpagam Academy of Higher Education Coimbatore India Department of Computer and Electronics Engineering SEST- Jamia Hamdard New Delhi India Department of Electronics and Communication Engineering KMEA College of Engineering Kerala Cochin India Department of Computer Science and Engineering V.S.B College of Engineering Technical Campus Coimbatore India

In recent years, transformer-based photo captioning frameworks plays a crucial role in improving individuals’ overall well-being, self-reliance, and inclusivity by giving them access to visual content via written and voiced explanations. This research investigates a two-layer transformer architecture to capture extensive relationships in visual data effectively. Using a two-layer transformer design, visual attention mechanism enables the model to autonomously identify relevant regions within a picture while generating a particular word within the anticipated caption. To get the features from the images, a pre-trained Inception V3 model is used, and the feature maps are extracted from the final convolutional block. The dataset used in this research was Flick8k, Flick30k, among bespoke datasets. The model was trained on a TPU platform and implemented on a Raspberry Pi 4B graphics processing unit. To assess the efficacy of the two-layer transformer model for photo captioning, a series of experiments were undertaken using benchmark datasets and with state-of-the-art models. The quality of the generated captions was evaluated using quantitative criteria such as BLEU, BLEUR, ROUGE and F1 score. Furthermore, a series of qualitative assessments were conducted, employing human evaluators, to measure the overall coherence, and fluency of the captions. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Long Term UGC Video Quality Evaluation Based on Transformer 24

Long Term UGC Video Quality Evaluation Based on Transformer

引用

3rd International Conference on computer, Artificial Intelligence and Control engineering, CAICE 2024

作者： Lang, Zhengang Zhang, Liyan School of Computer and Communication Engineering Dalian Jiaotong University China

ISBN: (纸本)9798400716942

User generated content (UGC) is gradually penetrating into people's lives, and how to improve the quality of UUC videos has been widely concerned, among which the quality evaluation of UGC videos is a key task. In this article, a novel video quality evaluation method is proposed, which utilizes video sampling technology to reduce computational complexity in the model inference process while maximizing the preservation of the original spatiotemporal information of the video. A novel temporal attention method has been proposed, specifically using 3D convolution and Transformer to weight the video frame scores by using long and short-term temporal information as attention weights. After systematic experimental verification, the proposed method in this article has significant effectiveness and excellent performance. © 2024 ACM.

关键词： Video analysis

来源：评论

学校读者我要写书评

暂无评论

Two-Dimensional Golay Complementary Array Sets with Arbitrary Lengths for Omnidirectional MIMO Transmission

引用

IEEE Transactions on communications 2024年第2期72卷 1135-1145页

作者： Zhao, You-Qi Pai, Cheng-Yu Huang, Zhen-Ming Liu, Zilong Chen, Chao-Yu National Cheng Kung University Department of Engineering Science Tainan701 Taiwan National Cheng Kung University Department of Engineering Science and the Institute of Computer and Communication Engineering Tainan701 Taiwan University of Essex School of Computer Science and Electronic Engineering ColchesterCO4 3SQ United Kingdom National Cheng Kung University Department of Electrical Engineering and the Institute of Computer and Communication Engineering Tainan701 Taiwan

This paper presents a coding approach for achieving omnidirectional transmission of certain common signals in massive multi-input multi-output (MIMO) networks such that the received power at any direction in a cell remains constant for any given distance. Specifically, two-dimensional (2D) Golay complementary array set (GCAS) can be used to design the massive MIMO precoding matrix so as to achieve omnidirectional transmission due to its complementary autocorrelation property. In this paper, novel constructions of new 2D GCASs with arbitrary array lengths are proposed. Our key idea is to carefully truncate the columns of certain larger arrays generated by 2D generalized Boolean functions. Finally, the power radiation patterns and numerical results are provided to verify the omnidirectional property of the GCAS-based precoding. The error performances of the proposed precoding scheme are presented to validate its superiority over the existing alternatives. © 1972-2012 IEEE.

关键词： Peak to average power ratio

来源：评论

学校读者我要写书评

暂无评论

Filtering Dense Dielectric Patch Antenna with a Pair of Step Arc-Shaped Slots

Filtering Dense Dielectric Patch Antenna with a Pair of Step...

引用

2024 International Conference on Intelligent communication, Sensing and Electromagnetics, ICSE 2024

作者： Sun, Wenzheng Wang, Hejie School of Computer and Communication Engineering Northeastern University Qinhuangdao China

ISBN: (纸本)9798350354935

This paper proposes a novel dense dielectric patch antenna (DDPA) with a flltering function, featuring an arc-shaped silver-plated patch. The antenna is composed of a dense dielectric patch (DDP) and a dual-layer PCB. A pair of stepped arc-shaped slots are designed on the DDP, which confines the resonant field region of the TE11 mode mainly to the area inside the stepped arc-shaped slot. As a result, the TE11 mode resonant frequency shifts to a higher frequency, forming a passband with the TE12 mode. The arc-shaped part in the middle of the DDP is silver-plated, altering the original electric field distribution and introducing a radiation zero at the lower frequency edge of the passband. By loading parallel crossed stubs onto the microstrip line, the interference in the high-frequency stopband is effectively suppressed. Therefore, this filtering antenna achieves a filtering function without requiring additional filtering circuits. Simulation results indicate that the antenna provides a relative bandwidth of 14.8% within the passband (0.58 GHz, 3.62 GHz∼4.2 GHz) with an average gain of 6.37 dBi. © 2024 IEEE.

关键词： Microstrip antennas

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：