检索结果-内蒙古大学图书馆

4th international conference on Frontiers technology of Information and Computer, ICFTIC 2022

作者： Xia, Chenyue Wang, Rui Deng, Zhuofu Zheng, Yingnan Software College of Northeastern University Liaoning Province Shenyang City China

ISBN: (纸本)9798350321951

With the rise of cloud computing services around the world, how to reduce the load balancing degree of cloud computing resource scheduling and improve the utilization rate of cloud computing resources has gradually attracted the attention of the academic community. In order to meet these needs, I proposed to use ant colony optimization algorithm scheduling to deal with large-scale cloud computing resource data sets. Meanwhile, in order to prevent the ant colony optimization algorithm I proposed from falling into the local optimal solution, I proposed to improve the probability and heuristic factor of selecting the next node of ant colony optimization algorithm as well as the update of pheromone, so as to give full play to the guiding role of pheromone to achieve the optimal cloud computing resource scheduling scheme. Examples have proved that the algorithm proposed by me has low relative standard deviation and load balancing degree. Meanwhile, the proposed scheduling scheme can evenly allocate resources, and the total utility value of scheduling is high, which can improve the scheduling ability of cloud computing resources to the greatest extent. At the same time, the fewer iterations required to schedule resources, the better the convergence of the algorithm. It can be seen that this algorithm can significantly improve the potential of cloud computing resource scheduling. © 2022 IEEE.

关键词： Cloud computing

来源：评论

学校读者我要写书评

暂无评论

DESIGN AND BENCHMARKING OF A LOW-COST RISC-V-BASED HIGH-PERFORMANCE computing CLUSTER FOR EDGE computing

DESIGN AND BENCHMARKING OF A LOW-COST RISC-V-BASED HIGH-PERF...

引用

2025 international conference on Energy, Power, Environment, Control and computing, ICEPECC 2025

作者： Hussain, Tassadaq Tahir, Muhammad Wasay Mushtaq, Manahil Khalid, Sidra Centre for Ai and Big Data Namal University Mianwali Pakistan Electrical Engineering Department Namal University Mianwali Pakistan UCERD PakASIC Pakistan Supercomputing Islamabad Pakistan

ISBN: (纸本)9781837243235

The exponential growth of IoT devices has led to a increase in data generation, necessitating real-time processing for critical decision-making, all while ensuring energy efficiency and privacy. Edge computing addresses these challenges by reducing latency through localized processing, but many edge devices still face limitations in real-time performance, energy efficiency, and scalability. In this work, we propose an energy-efficient, low-power HPC cluster based on single-board computers and open hardware. This cluster scales HPC capabilities with low-power, open-source solutions, making it ideal for IoT applications such as predictive maintenance in smart factories and real-time patient monitoring in healthcare. Our 10-node cluster, each featuring a StarFive JH7110 64-bit Quad-core CPU with an integrated GPU, achieves significant performance improvements. The computation time for a 1000x1000 matrix is reduced from 10.39 seconds on a single node to 1.637 seconds on 10 nodes, resulting in a 63% speedup. We introduce a custom distributed logistic regression algorithm designed for RISC-V, allowing model training on a cluster with improved scalability and performance. Our results show significant speedup in training time, with up to 5.6 times faster convergence as the number of nodes increases to 10. This work demonstrates the potential of RISC-V-based systems for efficient and distributed AI model training, making them well suited for decentralized applications in fields like academia, healthcare and finance while facilitating role of RISC-V in HPC. © The Institution of Engineering & technology 2025.

关键词： Edge computing

来源：评论

学校读者我要写书评

暂无评论

An improved Harris Hawk optimization algorithm and its application to Extreme Learning Machine 3

An improved Harris Hawk optimization algorithm and its appli...

引用

3rd international conference on Consumer Electronics and Computer Engineering, ICCECE 2023

作者： Liu, Ziliang Chen, Hongwei School of Computer Science Hubei University of Technology Wuhan China

ISBN: (纸本)9798350331578

The Harris Hawk optimization (HHO) algorithm is an excellent swarm intelligence optimization algorithm which has the advantages of high efficiency in finding the best, ease of implementation and wide application. It also has some disadvantages such as the possibility of convergence too fast and the tendency to fall into local optima. This paper combines an improved escape energy update approach and the leader update operator of the Salp Swarm Algorithm to improve the HHO, named IMHHO. The experiments show that the improvements have improved the algorithm's ability to find the best. IMHHO was also used in the parameter optimization of the Extreme Learning Machine, which also enables the ELM to find the right weights and bias values and to regress the data more accurately. © 2023 IEEE.

关键词： Optimization

来源：评论

学校读者我要写书评

暂无评论

On the ICN-IoT with federated learning integration of communication: Concepts, security-privacy issues, applications, and future perspectives

引用

FUTURE GENERATION COMPUTER SYSTEMS-THE international JOURNAL OF ESCIENCE 2023年 138卷 61-88页

作者： Rahman, Anichur Hasan, Kamrul Kundu, Dipanjali Islam, Md. Jahidul Debnath, Tanoy Band, Shahab S. Kumar, Neeraj Mawlana Bhashani Sci & Technol Univ Dept CSE Tangail Bangladesh Constituent Inst Univ Dhaka Natl Inst Text Engn & Res NITER Dept CSE Savar Dhaka 1350 Bangladesh Hankuk Univ Foreign Studies Dept ICE Seoul South Korea Green Univ Bangladesh Dept CSE 220-D Begum Rokeya Sarani Dhaka 1207 Bangladesh Natl Yunlin Univ Sci & Technol Coll Future Future Technol Res Ctr 123 Univ Rd Sect 3 Touliu 64002 Yunlin Taiwan Deemed Univ Thapar Inst Engn & Technol Dept CSE Patiala Punjab India Univ Petr & Energy Studies Sch Comp Sci Dehra Dun Uttarakhand India Asia Univ Dept Comp Sci & Informat Engn Taichung Taiwan King Abdulaziz Univ Jeddah Saudi Arabia

The individual and integration use of the Internet of Things (IoT), Information-Centric Networking (ICN), and Federated Learning (FL) have recently been used in several network-related scenarios and have consequently experienced a growing interest in the research community. Federated learning addresses the privacy and security issues of the IoT data in a decentralized manner. Also, it can be capable of training the multiple learning algorithms through local content except for exchanging data through intelligent Artificial Intelligence (AI)-based algorithms. Moreover, in ICN, the content is retrieved and stored based on the content name rather than the content location address. On the other hand, it is challenging to support the massive IoT devices by the fifth generation (5G) mobile-cellular networks. Therefore, the cellular 6G networks are expected to increase the connection capabilities by 10-100 times over 5G, which necessitates a convergence of Communication, computing, and Caching (3C). At the same time, the in-network caching capabilities of ICN can be attractive features for IoT networks. IoT aspires to link anybody and/or everything at any time and location. However, integrating IoT with different areas is a new academic topic and is still in its infancy. As a result, this research highlights the potential of ICN for IoTs by conducting an exhaustive literature review. This work provides a comprehensive survey regarding these three recent research trends (i.e., FL, IoT, and ICN) and reviews the related state-of-the-art literature. We first describe the main features of each technology and discuss their most common and used variants. Furthermore, we envision the integration of such technologies to take advantage efficiently. Indeed, we consider their group-wise (FL-ICN-IoT) utilization based on the need for more robust security and privacy. Additionally, we cover the application fields of these technologies both individually and combinedly. Finally, we discuss

关键词： Internet of Things Information-Centric Networking Federated Learning Communication Security computing Machine Learning Privacy Artificial Intelligence Confidentiality

来源：评论

学校读者我要写书评

暂无评论

Joint computing Offloading and Trajectory for Multi-UAV Enabled MEC Systems 22

Joint Computing Offloading and Trajectory for Multi-UAV Enab...

引用

22nd IEEE international conference on Communication technology, ICCT 2022

作者： Xu, Wenlong Zhang, Tiankui Yang, Liwei School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing China School of Information and Electrical Engineering China Agricultural University Beijing China

ISBN: (纸本)9781665470674

The cooperation of multiple unmanned aerial vehicles (UAVs) is investigated to provide auxiliary computing services for ground users. First, the system cost is defined considering the energy consumption of the UAV, the energy consumption of the user, and the delay of the user at the same time. Take into account the dynamic allocation of bandwidth by the user and the dynamic allocation of computational resources by the UAV, the flight trajectory of UAVs, the offloading object and the offloading ratio of users are jointly optimized to minimize the system cost. Due to the dynamic and long-term feature of the problem, it is described as a Markov decision process. A joint computing offloading and trajectory algorithm is proposed based on the PPO in deep reinforcement learning. Simulation results show the convergence of the proposed algorithm. The proposed algorithm has superior performance compared with the benchmark algorithms. © 2022 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Application of Deep Learning Model Inference with Batch Size Adjustment 13

Application of Deep Learning Model Inference with Batch Size...

引用

13th international conference on Information and Communication technology convergence, ICTC 2022

作者： Oh, Seungtaek Moon, Jaewon Kum, Seungwoo Information Media Research Center Korea Electronics Technology Institute Seoul Korea Republic of

ISBN: (数字)9781665499392

ISBN: (纸本)9781665499392

The demand for deep learning model applications by users on edge devices is increasing. If the AI model inferences are directly operated on the edge device rather than the cloud or server, it allows customers to proceed additional tasks more conveniently. In order for a deep learning model to be used more efficiently in an embedded environment, studies such as model weight reduction and network compression have been conducted. In addition, studies related to GPU optimization that maximize the usage of limited GPU resources are also attracting a lot of attention. In this paper, in order to maximize the GPU allocation of edge resources, we conducted some experiments to infer images of multiple frames at once, not just one input. For the experiments, the YOLOv4 model was converted into several models with different input layer shapes, and the inference was performed by adjusting the input batch size. NVIDIA Jetson AGX Xavier was used as the edge device, and all other models with different batch sizes obtained the same inference results. As the number of batches increases as much as possible according to the resource limit, the inference speed increases. © 2022 IEEE.

关键词： Edge computing

来源：评论

学校读者我要写书评

暂无评论

Optimizing Supervised Learning of Deep Spiking Neural Network towards Memristor Crossbar Implementation 15

Optimizing Supervised Learning of Deep Spiking Neural Networ...

引用

15th IEEE international conference on ASIC, ASICON 2023

作者： Chen, Qi Zhang, Dayou Fu, Jiawei He, Yuhui Huazhong University of Science and Technology School of Integrated Circuits Wuhan China

ISBN: (纸本)9798350312980

The gradient-based direct training with spike timing coding known as one major approach to supervised learning of deep spiking neural network often encounters problems like failure of convergence and quite low training efficiency. Here we develop a novel training framework by incorporating several algorithm-hardware codesign strategies: first, we optimize the expression of synaptic weight update by rewriting it as a product of the forward signal and backward error, and thus the expression becomes as convenient as that of conventional backpropagation for hardware implementation;second, coerced firing is exerted on those silent neurons to circumvent the problem of gradient exploding;finally, a stochastic computing technique is employed to formulate the write pulse sequence so that the memristor-based synaptic array can be updated in a one-step manner. Our network level simulation results show >90% accuracy on MNIST benchmark while with significantly enhanced training efficiency. © 2023 IEEE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Ant Colony Approach with Subjective-Objective Weight Optimization for Service Selection 19

An Ant Colony Approach with Subjective-Objective Weight Opti...

引用

2022 IEEE SmartWorld, 19th IEEE international conference on Ubiquitous Intelligence and computing, 2022 IEEE international conference on Autonomous and Trusted Vehicles conference, 22nd IEEE international conference on Scalable computing and Communications, 2022 IEEE international conference on Digital Twin, 8th IEEE international conference on Privacy computing and 2022 IEEE international conference on Metaverse, SmartWorld/UIC/ATC/ScalCom/DigitalTwin/PriComp/Metaverse 2022

作者： Yan, Xiukun Li, Jing Zhu, Ming Du, Jiayi Xie, Yong Zhao, Lianjun Shandong University of Technology College of Computer Science and Technology Zibo China Central South University of Forestry and Technology College of Computer and Information Changsha China Nanjing University of Posts and Telecommunications Department of Computer Science Nanjing China

ISBN: (纸本)9798350346558

To solve the problem that current service selection algorithms converge slowly and cannot simultaneously consider user requirements as well as service performance, an ant colony approach with subject-objective weight optimization for service selection is proposed. First, we use the subjective and objective weight analysis method to preprocess services and convert each service's multiple Quality of Service attribute into a normalized value. Then, three optimizations are proposed for the original ant colony optimization algorithm. The initial pheromone concentration is unevenly distributed according to the integrated value, the influencing factors of the algorithm are changed from fixed values to dynamic adaptive adjustment values, the pheromone update rule is improved. Finally, an enhanced ant colony optimization algorithm is applied to service selection. Experimental results show that the proposed method has a faster convergence rate than other referencing algorithms in dealing with the service selection problem, and the selection result satisfies the users better. © 2022 IEEE.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

Deep reinforcement learning based delay-Aware task offloading for UAV-Assisted edge computing 5

Deep reinforcement learning based delay-Aware task offloadin...

引用

5th international conference on Algorithms, computing and Artificial Intelligence, ACAI 2022

作者： Huang, Mingxuan Sun, Kaixuan Hou, Yunpeng Ye, Zhicheng Wan, Yuanlong He, Huasen University of Science and Technology of China Hefei China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Hefei China

ISBN: (纸本)9781450398343

Multi-Access edge computing has been widely used in various Internet of Things (IoT) devices because of its excellent computing and fast interaction abilities. How to improve the service extensibility of edge computing and optimize the computing offload strategy has become the key to improve the quality of service to edge computing users. However, the traditional offloading strategy based on mathematical programming has exposed its inherent limitations in dynamic scenarios, and cannot meet the requirements of multiple-mobile terminals distributed in a large area. Therefore, this paper use Unmanned Aerial Vehicles (UAVs) to establish a multi-UAV-Assisted edge computing framework for extending the service range, and proposes an offloading strategy based on reinforcement learning to offload the growing computing requirements from mobile terminals to edge servers. By mapping the states of mobile terminals and UAVs to the corresponding action space, and then offloading computing tasks to UAVs, the energy consumption caused by computing and processing tasks of mobile terminals can be effectively reduced. Jointly considering the potential dimensional disaster of state space and the convergence failure imposed by the increase of device numbers, a novel computation offloading strategy based on deep reinforcement learning is proposed. Moreover, we design a load-balancing mechanism in the UAVs to improve the processing capacity. Experimental results prove that our proposed algorithm can effectively reduce the computing energy consumption of mobile terminals and avoid task timeout with a short convergence time. © 2022 ACM.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

Research on lightweight detection algorithm of Infrared Image based on YOLOv5 4

Research on lightweight detection algorithm of Infrared Imag...

引用

4th international conference on Advanced Algorithms and Signal Image Processing, AASIP 2024

作者： Zhang, Wenyi Mo, Hongwei Wu, Songling Li, Xiaomin Harbin Engineering University School of Intelligent Science and Engineering Harbin150000 China Beijing Aerospace Automatic Control Institute Beijing100854 China National Key Laboratory of Science and Technology on Aerospace Intelligence Control Beijing100854 China

ISBN: (数字)9781510682818

ISBN: (纸本)9781510682801

At present, infrared image can show good detection effect in complex environment because of its unique imaging characteristics, which brings a lot of convenience to people's production and life. However, because the infrared image contains a great lot of unnecessary information, it has great requirements for computing power and memory capacity, so it is challenging to satisfy the demands of real-time detection on resource-constrained portable mobile devices. To solve this problem, this paper proposes a refined lightweight algorithm built upon the YOLOv5, which changes the backbone network of the model to GhostNet network, introduces CA attention mechanism into the model to improve the detection effect of small targets, and changes the loss function of the model to Focal-EIoU loss function to improve the location effect of the boundary box and quicken the model's rate of convergence. Then replace the conventional convolution used in the sampling process of the model with the space-to-depth convolution module, in order to enhance the model's detection performance on low-resolution images and enable the extraction of more feature information. According to the experimental results, the enhanced detection algorithm is just 1% less than the initial YOLOv5 algorithm, and the model volume and floating point calculation are reduced by 31.9% and 36% respectively. This is better suited for deployments that are lightweight. © 2024 SPIE.

关键词： Convolution

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：