检索结果-内蒙古大学图书馆

2024 IEEE Conference on Network Function Virtualization and Software Defined Networks, NFV-SDN 2024

作者： Das, Debojeet Lakshya Vutukuru, Mythili Department of Computer Science and Engineering Indian Institute of Technology Bombay India

ISBN: (纸本)9798350380538

To leverage the full potential of high-speed network interface cards (NICs), it is essential for network applications to use high-performance packet I/O frameworks. However, porting applications to use these specialized frameworks is a tedious task. While prior work has proposed frameworks to ease the development of network functions over fast packet I/O mechanisms, there has not been much work done to ease the porting of applications that are built to run on the kernel network stack. This paper presents Fastlane, a framework to easily port the "fast path" of traditional network applications to use fast packet I/O mechanisms while retaining the "slow path" to run on the traditional network stack. Fastlane provides APIs by which application developers can configure, use, and switch between different fast packet I/O mechanisms like DPDK and CNDP in the application’s fast path. Additionally, Fastlane provides APIs for the fast path to communicate with the slow path running on the kernel network stack. Fastlane also allows applications to be deployed in an unprivileged Kubernetes pod, easing application development for cloud deployments. We evaluate the effectiveness of Fastlane by porting a production-grade 5G user plane function to use our framework. Our evaluation shows that Fastlane enables developers to build high-performance network applications easily and imposes minimal performance overheads. ©2024 IEEE.

关键词： HIgh speed networks

来源：评论

学校读者我要写书评

暂无评论

MORE-Router+: Multilayer Multi-capacity ORdered Escape Routing via Bus-oriented Layer Assignment 25

MORE-Router+: Multilayer Multi-capacity ORdered Escape Routi...

引用

25th International Symposium on Quality Electronic Design, ISQED 2024

作者： Gao, Zhenyi Dong, Sheqin Cheng, Zifei Yu, Wenjian BNRist Tsinghua University Department of Computer Science & Technology Beijing China

ISBN: (纸本)9798350309270

Escape routing is one of the fundamental problems in advanced packaging and PCB design. In this work, the multilayer multi-capacity ordered escape routing on grid pin array (GPA) is considered with a bus-oriented layer assignment. The nets are grouped as a set of buses and the problem objective is minimizing the number of routing layers while satisfying the layer-adjacency, capacity and projection ordering constraints. We propose a router for multilayer multi-capacity ordered escape routing (named MORE-Router+), which is based on integer linear programming (ILP) and consists of bus planning, layer assignment and detailed routing. In the bus planning stage we assign one feasible escape direction for each bus. The layer assignment algorithm determines the layer location so that the number of used layers is minimized and the layer-adjacency constraint is satisfied. For the detailed routing, we extend the multi-capacity min-cost multi-commodity flow (MC-MCF) graph [1] to multilayer MC-MCF graph and formulate the problem as an ILP problem. Furthermore, a net sequence division technique is proposed to accelerate the detailed routing. The proposed method is tested on 9 cases with up to 54 buses and 1602 nets in 50 × 50 GPA. The results show that the net sequence division technique brings 4.06X through over 156X speedup, and MORE-Router+ costs no more than 266 seconds to complete the multilayer multi-capacity ordered escape routing. © 2024 IEEE.

关键词： Polychlorinated biphenyls

来源：评论

学校读者我要写书评

暂无评论

Multi-Medium Handover Transactions with Interactive and Protected Self-Intelligence Principles for Unified Military Vehicular Sensor Networks 3

Multi-Medium Handover Transactions with Interactive and Prot...

引用

3rd International Conference on Sustainable Mobility Applications, Renewables and technology, SMART 2024

作者： Rajasoundaran, Soundararajan Galety, Mohammad Gouse Natarajan, Arul Kumar Diyorbek, Gaffarov Umedjon, Sattorov Department of Computer Science Samarkand International University of Technology Samarkand Uzbekistan

ISBN: (纸本)9798331532925

Unified Military Vehicular Sensor Network (UM-VSNET) aims to build a multi-base vehicular sensor network using aerial, land-based, and underwater sensor nodes under a single architecture. The future VSNET used in military environments requires secure and unified control optimization strategies at various war bases such as land, air, sea surface, and undersea. The unmanned army vehicles used in land, air, sea, and undersea environments get more crucial spots to counter any vulnerabilities and provide sustainable electronic battlefield security features. The recent works in this domain have limitations on heterogeneous and multi-medium handover opportunities. Most of the existing mechanisms mainly execute uniform handover protocols among wireless nodes. These are insufficient for unified military vehicular networks since they are more versatile regarding medium principles, internal resources, functional targets, security expectations, and autonomous capabilities. On the scope, the proposed model initiates heterogeneous handover principles with Artificial Intelligence (AI) and interactive autonomous capabilities. The proposed research creates Unmanned Aerial Vehicles (UAV), terrestrial mobile, and maritime network models with crucial configuration parameters. The article proposes Multi-Medium Handover Transactions with Interactive and Protected Self-Intelligence Principles for Unified Vehicular Sensor Networks (MH-IPUV) on the developed network platform. MH-IPUV consists of heterogeneous medium access control policies, vehicular mobility models (aerial, land, and sea), distributed confidential (Advanced Encryption Standard, AES) and authentication procedures (Elliptic Curve Cryptography-Digital Signature, ECCDS), Secure Hashing Algorithm (SHA-3), and secure AI-based handover principles. In the proposed military network environment, each node (UAVs, Terrestrial Nodes, and Maritime Nodes) has been implemented with the algorithms of the MH-IPUV system. Considering the sec

关键词： Sensor nodes

来源：评论

学校读者我要写书评

暂无评论

Depression Detection from Text Classification Using Neural Network 8

Depression Detection from Text Classification Using Neural N...

引用

8th International Conference on Information technology, Information Systems and Electrical Engineering, ICITISEE 2024

作者： Fabian Sutrisno, Neil Errando Ebenezer, Joseph Adiel Edbert, Ivan Sebastian Philip, Samuel School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350368970

Depression has become a common mental health problem in our society. In fact, according to WHO, depression is a mental health that occurs to 10% of the global population every year. The risk of failure to handle depression is the risk of suicide. People often use social media to express their feelings. Artificial Intelligence, one technological improvement, was created to solve cognitive problems. One of the capabilities to interpret and predict results from text is called Natural Language Processing (NLP). In this research, the researchers propose the idea of forming a model that can classify whether the text entered is classified as depression or non-depression. This research focuses on using deep learning as a model, ratio comparisons in the dataset, and using a bidirectional layer for the models. After conducting the study, the best model is the Bi-GRU model with the dataset ratio of 80:10:10, where this model achieves an accuracy of 0.9955 and a loss accuracy of 0.0154. Hopefully, this model will continue this research until it impacts the surrounding environment. © 2024 IEEE.

关键词： Bidirectional Deep Learning Depression GRU LSTM Text Classification

来源：评论

学校读者我要写书评

暂无评论

Applications and Associated Challenges in Deployment of Software Defined Networking (SDN)

Applications and Associated Challenges in Deployment of Soft...

引用

International Conference on Emerging Trends and Applications in Artificial Intelligence, ICETAI 2023

作者： Baniya, Pashupati Agrawal, Atul Nand, Parma Bhushan, Bharat Hameed, Alaa Ali Jamil, Akhtar Department of Computer Science and Engineering School of Engineering and Technology Sharda University Greater Noida India Department of Computer Engineering Istinye University Istanbul Turkey Department of Computer Science National University of Computer and Emerging Sciences Islamabad Pakistan

ISBN: (纸本)9783031567278

SDN, a rising technology within the realm of Internet of Things (IoT), has been increasingly well-received in recent times. This article presents a summary of SDN along with its different elements, advantages, and difficulties. The paper aims to provide practical solutions for introducing OpenFlow into commercial routers without hardware modifications and extending the integration of OpenFlow with legacy control protocols and control planes. In addition, the paper presents a refactoring process for migrating traditional network applications to OpenFlow-based ones, focusing on the security challenges and techniques of open technologies like SDN, OpenROADM, and SDN-based Mobile Networks (SDMN). The document also examines the advantages and possible uses of SDMN in enhancing network adaptability, streamlining network administration, and bolstering network security. The article also discusses O-RAN network innovations and difficulties, such as AI and ML workflows that are made possible by the architecture and interfaces, security concerns, and, most importantly, standardization issues. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Software defined networking

来源：评论

学校读者我要写书评

暂无评论

IoMT enabled diabetic retinopathy segmentation and classification using ensemble efficient net model

引用

Multimedia Tools and Applications 2025年第18期84卷 20047-20078页

作者： Bhutnal, Vinodkumar Moparthi, Nageswara Rao Department of Computer Science and Engineering Koneru Lakshmaiah Education Foundation Andhra Pradesh Vaddeswaram India

The most common preventable cause of blindness in working-age adults worldwide is diabetic retinopathy (DR). Accurate detection of DR by machine learning (ML) approaches is generally limited to pre-selected features. However, the feature learning ability and parameter consideration must be efficient to achieve better results. Therefore, this research proposes a novel automation of DR using images in the Internet of Medical Things (IoMT) established on a deep learning (DL) model to improve segmentation and classification performance. In this work, IoMT datasets such as the Diabetic Retinopathy Dataset (DRR) and the Indian Diabetic Retinopathy Image Dataset (IDRiD) are collected from public sources. First, the collected fundus images are pre-processed using the Sequential Bilateral-Rolling Guidance Filter (SeQ-BR) for noise exclusion. The filtered images are subjected to the Improved Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm to improve contrast. The pre-processed images are segmented into thin vessels, exudates, orbit, microaneurysms, and optic nerve heads using the Dilatation U-Net model (Di-UNet). From the segmented images, the appropriate features are extracted using the Convolutional Swin Transformer (CST) model. To minimize the dimensionality problems, an Opposition Based Mud Ring Optimization (OMR) algorithm is used to select the features. Finally, disease severity can be classified using the ensemble efficient net-assisted gazelle optimization algorithm model (E-EffGaz). The proposed method results in 97.69% accuracy for the DRR dataset and 97.52% accuracy using the IDRiD dataset. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Comparative Analysis of Machine Learning Algorithms for Laptop Value Estimation 9th

Comparative Analysis of Machine Learning Algorithms for Lapt...

引用

9th International Conference on computer science and Computational Intelligence, ICCSCI 2024

作者： Hansel, Kent Christopher Tanoto, Vincentius Axelle Suri, Puti Andam Fajar, Muhamad Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia

The increasing prevalence of laptops in homes, schools, and workplaces makes accurate value estimations essential to guide buyers in making informed decisions. This study compares the effectiveness of four machine learning algorithms-Multiple Linear Regression (MLR), Artificial Neural Network (ANN), Random Forest (RF), and Support Vector Machine (SVM)-in predicting laptop prices based on specifications such as brand, processor speed, RAM size, storage capacity, screen size, and weight. The performance of these models is evaluated using Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Root Mean Squared Error (RMSE), and R-squared (R2) metrics. Results indicate that MLR and RF provide superior performance, achieving the lowest RMSE of 0.02 and the highest R2 score of 1.0, demonstrating their suitability for datasets with linear feature relationships. The findings highlight the significance of selecting appropriate models based on the nature of the dataset, suggesting that while ANN and SVM are potent for complex, non-linear data, MLR and RF are preferable for linear datasets. © 2024 The Authors.

关键词： Multiple linear regression

来源：评论

学校读者我要写书评

暂无评论

LSTM-based Framework for 5G Resource Allocation Prediction 11

LSTM-based Framework for 5G Resource Allocation Prediction

引用

11th International Symposium on Telecommunication, IST 2024

作者： Pourmahboubi, Amin Tabrizchi, Hamed University of Tabriz Faculty of Mathematics Statistics and Computer Science Department of Computer Science Tabriz Iran

ISBN: (纸本)9798350356250

During the rapid evolution of 5G networks, the efficient allocation of bandwidth, frequency spectrum, and computing power is critical to maintaining a high standard of service and performance. A methodology for predicting resource allocation percentages based on Long-Short-Term Memory (LSTM) is presented in this paper. The ability to predict resource allocation percentages enhances network operators' efficiency in terms of bandwidth, frequency spectrum, and computing power utilization. Network operators can maintain high service levels by accurately predicting resource needs and allocating resources dynamically. In high-latency and high-reliability applications such as VoIP calls, video streaming, and emergency services, this is particularly critical. From Iot sensor data to high-definition video calls and emergency services, our proposed model explores the specific resource requirements of a 5 G network. Our experimental results show that the LSTM model achieves a considerable improvement in performance when compared to the best-performing machine learning model, with an R2 score of 0.976 compared to 0.937 for the best-performing model. Predictive accuracy is improved by approximately 4.16 percent. This highlights the superiority of our proposed framework to recognizing patterns in time series data. © 2024 IEEE.

关键词： 5G mobile communication systems

来源：评论

学校读者我要写书评

暂无评论

AI-Generated Content-as-a-Service in IoMT-Based Smart Homes: Personalizing Patient Care With Human Digital Twins

引用

IEEE Transactions on Consumer Electronics 2024年第1期71卷 1-1页

作者： Akram, Junaid Aamir, Misha Raut, Roshani Anaissi, Ali Jhaveri, Rutvij H. Akram, Awais School of Computer Science The University of Sydney Camperdown NSW Australia Department of Information Technology Pimpri Chinchwad College of Engineering Savitribai Phule Pune University Pune India Department of Computer Science and Engineering School of Technology Pandit Deendayal Energy University India COMSATS University Islamabad Vehari Pakistan

We introduce the AI-Generated Optimal Decision (AIGOD) algorithm and the Deep Diffusion Soft Actor-Critic (DDSAC) framework, marking a significant advancement in integrating Human Digital Twins (HDTs) with AI-Generated Content (AIGC) within IoMT-based smart homes. Our innovative AI-Generated Content-as-a-Service (AIGCaaS) architecture, optimized for IoMT environments, leverages network edge servers to enhance the selection of AI-Generated Content Service Providers (AISPs) tailored to the unique characteristics of individual HDTs. Extensive experiments demonstrate DDSAC’s HDT-centric approach outperforms traditional Deep Reinforcement Learning algorithms, offering optimal AIGC services for diverse healthcare needs. Specifically, DDSAC achieved a 20% improvement in task completion rates and a 15% increase in overall utility compared to existing methods. These findings highlight the potential of HDTs in personalized healthcare by simulating and predicting patient-specific medical outcomes, leading to proactive and timely interventions. This integration facilitates personalized healthcare, establishing a new standard for patient-centric care in smart home environments. By leveraging cutting-edge AI techniques, our research significantly contributes to the fields of IoMT and AIGC, paving the way for smarter and more responsive healthcare services. IEEE

关键词： Real time systems

来源：评论

学校读者我要写书评

暂无评论

A Machine Learning-Based Attendance Management System

A Machine Learning-Based Attendance Management System

引用

2024 IEEE International Conference on Information technology, Electronics and Intelligent Communication Systems, ICITEICS 2024

作者： Mageshvaran, G. Sowmya, V.J. Hindustan Institute of Technology and Science Department of Computer Applications Chennai India

ISBN: (纸本)9798350382693

The primary goal of this project is to update and modernize the current system, as well as boost its efficacy and efficiency, by developing an attendance tracking system for educational institutions based on facial recognition technology. Recording and manually inputting attendance logs in a classroom is not the most efficient method. In the current age, some students may find it interesting or creative to skip classes or arrange for proxies for those who are absent. Manually recording attendance in logbooks becomes a challenging and readily manipulable activity. This technology uses facial recognition to automatically identify students in the classroom and record their attendance. In order to construct this system, real-time facial recognition is used. The reference faces in the dataset are compared with the detected faces, marking the participants' attendance The data is trained using the Fisher Face Recognizer algorithm, and faces are detected using the haar-cascade classifier. SMTP is utilized for email transmission and reception, and the Tkinter graphical user interface is employed for input capture. The technology in question is a facial recognition system. One of a person's natural features that can be used to identify them specifically is their face. Faces are used to trace identification since they are unlikely to deviate or be copied. The project comprises creating facial databases in order to enhance the recognizer algorithm's capabilities. During attendance sessions, a systematic comparison of faces against the established database will be conducted to accurately determine individual identities. Upon recognition, the presence of the individual will be automatically documented, and the pertinent information will be saved in an Excel spreadsheet. The excel document with every person's attendance detail is mailed to the relevant professors at the conclusion of the day. © 2024 IEEE.

关键词： Students

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：