检索结果-内蒙古大学图书馆

international conference on Computer and Communications (ICCC)

作者： Enchang Sun Mengsi Li Xiaoxuan Dong Dongying Zhang Faculty of Information Technology Beijing University of Technology Beijing China Beijing Dublin International College Beijing University of Technology Beijing China Information Technology Support Center Beijing University of Technology Beijing China

Multi-access edge computing (MEC) technology is a potential solution to the conflict between resource-constrained unmanned aerial vehicles (UAVs) and various emerging computationally intensive, delay-sensitive applications of UAVs. In this paper, we study the joint optimization problem of distributed computation offloading and resource allocation in MEC-supported UAV application scenarios, where the task of each UAV can be divided and offloaded to multiple servers deployed in ground wireless access points (WAPs) for parallel processing. To enhance the quality of user experience (QoE) while avoiding excessive optimization, we propose a new user satisfaction function to evaluate the task processing level of MEC network, which respectively scores the actual task processing delay and energy consumption based on different criteria determined by task requirements. We aim to trade off delay and energy consumption by jointly optimizing server selection, computing resource allocation, transmission power control, and task allocation to achieve maximum user satisfaction. The joint optimization problem is formulated as a mixed-integer nonlinear programming (MINLP) problem. To reduce the complexity of the joint optimization problem, it is decomposed into three sub-problems based on the divide-and-conquer approach: server computing resource allocation, transmission power control, and task offloading decision. We propose a three-layer intelligent optimization (TIO) algorithm, which employs the Lagrange multiplier method, the AdaDelta gradient descent method, and our proposed multivariate decoupling-particle swarm optimization (MD-PSO) algorithm to solve the above three sub-problems alternately for gradually approaching the maximum user satisfaction. The numerical results show that our proposed algorithm is significantly superior to the benchmark algorithms in terms of solving speed, accuracy, and stability.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Future generation of service-oriented computing systems

引用

FUTURE GENERATION COMPUTER SYSTEMS-THE international JOURNAL OF ESCIENCE 2021年 118卷 252-256页

作者： Yangui, Sami Goscinski, Andrzej Drira, Khalil Tari, Zahir Benslimane, Djamal Univ Toulouse LAAS CNRS INSA F-31400 Toulouse France Deakin Univ Geelong Vic Australia Univ Toulouse LAAS CNRS F-31400 Toulouse France RMIT Univ Sch Comp Technol Sydney NSW Australia Univ Lyon 1 LIRIS Lyon France

Service-Oriented computing (SOC) and SOC systems have been invented, studied, and developed in response to the problems generated by heterogeneity and poor latency, and a need for modularization and standardization. Wide application of SOC, practically in all areas of human endeavors, interconnectivity provided by wireless networks, huge volumes and diversity of data, increased frequency of security attacks, and reliability expectations have increased research, development, and deployment of SOC systems. These circumstances and requirements could only be dealt with by a future generation of SOC systems. This implies a need for fast and high-quality research and wide and efficient dissemination of achieved innovative results. Therefore, a Special Issue (SI) was proposed, and its relevant scope defined. This SI involves a set of high-quality and cross-community scientific papers from various disciplines (e.g., cloud computing, edge computing, business process management). The accepted papers focus on research leading toward future generation of service-oriented computing systems (e.g., dynamic quality of service management in fog computing, time-aware resources allocation in cloud computing, resilient composition in the Internet of Things) with emphasis on advanced results that solve open research problems and have significant impact on the field of service-oriented computing. (C) 2021 Published by Elsevier B.V.

关键词： Service computing Cloud computing Fog computing Business Process Management

来源：评论

学校读者我要写书评

暂无评论

36th international conference on High Performance computing, ISC High Performance 2021

36th International Conference on High Performance Computing,...

引用

36th international conference on High Performance computing, ISC High Performance 2021

ISBN: (纸本)9783030787127

The proceedings contain 24 papers. The special focus in this conference is on High Performance computing. The topics include: Proctor: A Semi-Supervised Performance Anomaly Diagnosis Framework for Production HPC Systems;COSTA: Communication-Optimal Shuffle and Transpose Algorithm with Process Relabeling;Enabling AI-Accelerated Multiscale Modeling of Thrombogenesis at Millisecond and Molecular Resolutions on Supercomputers;Evaluation of the NEC Vector Engine for Legacy CFD Codes;distributed Sparse Block Grids on GPUs;iPUG: Accelerating Breadth-First Graph Traversals Using Manycore Graphcore IPUs;Optimizing GPU-Enhanced HPC System and Cloud Procurements for Scientific Workloads;a Performance Analysis of Modern parallel Programming Models Using a Compute-Bound Application;analytic Modeling of Idle Waves in parallel Programs: Communication, Cluster Topology, and Noise Impact;BluesMPI: Efficient MPI Non-blocking Alltoall Offloading Designs on Modern BlueField Smart NICs;performance of the Supercomputer Fugaku for Breadth-First Search in Graph500 Benchmark;Under the Hood of SYCL – An Initial Performance Analysis with An Unstructured-Mesh CFD Application;Characterizing Containerized HPC Applications Performance at Petascale on CPU and GPU Architectures;ubiquitous Performance Analysis;artemis: Automatic Runtime Tuning of parallel Execution Parameters Using Machine Learning;Lessons Learned from Accelerating Quicksilver on Programmable Integrated Unified Memory Architecture (PIUMA) and How That’s Different from CPU;a Hierarchical Task Scheduler for Heterogeneous computing;auto-Precision Scaling for distributed Deep Learning;FPGA Acceleration of Number Theoretic Transform;Designing a ROCm-Aware MPI Library for AMD GPUs: Early Experiences;A Tunable Implementation of Quality-of-Service Classes for HPC networks;scalability of Streaming Anomaly Detection in an Unbounded Key Space Using Migrating Threads.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Adaptive distributed Convolutional Neural Network Inference at the Network Edge with ADCNN 20

Adaptive Distributed Convolutional Neural Network Inference ...

引用

49th international conference on parallel Processing (ICPP)

作者： Zhang, Sai Qian Lin, Jieyu Zhang, Qi Harvard Univ Cambridge MA 02138 USA Univ Toronto Toronto ON Canada Microsoft Redmond WA USA

ISBN: (纸本)9781450388160

The emergence of the Internet of Things (IoT) has led to a remarkable increase in the volume of data generated at the network edge. In order to support real-time smart IoT applications, massive amounts of data generated from edge devices need to be processed using methods such as deep neural networks (DNNs) with low latency. To improve application performance and minimize resource cost, enterprises have begun to adopt Edge computing, a computation paradigm that advocates processing input data locally at the network edge. However, as edge nodes are often resource-constrained, running data-intensive DNN inference tasks on each individual edge node often incurs high latency, which seriously limits the practicality and effectiveness of this model. In this paper, we study the problem of distributed execution of inference tasks on edge clusters for Convolutional Neural networks (CNNs), one of the most prominent models of DNN. Unlike previous work, we present Fully Decomposable Spatial Partition (FDSP), which naturally supports resource heterogeneity and dynamicity in edge computing environments. We then present a compression technique that further reduces network communication overhead. Our system, called ADCNN, provides up to 2.8x speed up compared to state-of-the-art approaches, while achieving a competitive inference accuracy.

关键词： Edge computing Convolutional neural networks distributed inference

来源：评论

学校读者我要写书评

暂无评论

Single Node Acceleration of Generative Adversarial networks using HPC for Image Analytics 22

Single Node Acceleration of Generative Adversarial Networks ...

引用

Proceedings of the 2022 5th international conference on Computational Intelligence and Intelligent Systems

作者： Aswathy Ravikumar Harini Sriraman School of Computer Science and Engineering Vellore Institute of Technology India

ISBN: (纸本)9781450397612

Generative Adversarial networks (GAN) are approaches that are utilized for data augmentation, which facilitates the development of more accurate detection models for unusual or unbalanced datasets. Computer-assisted diagnostic methods may be made more reliable by using synthetic pictures generated by GAN. Generative adversarial networks are challenging to train because too unpredictable training dynamics may occur throughout the learning process, such as model collapse and vanishing gradients. For accurate and faster results the GAN network need to trained in parallel and distributed manner. We enhance the speed and precision of the Deep Convolutional Generative Adversarial networks (DCGAN) architecture by using its parallelism and executing it on High-Performance computing platforms. The effective analysis of a DCGAN in Graphic Processing Unit and Tensor Processing Unit platforms in which each layer execution pattern is analyzed. The bottleneck is identified for the GAN structure for each execution platforms. The Central Processing Unit is capable of processing neural network models, but it requires a great deal of time to do it. Graphic Processing Unit in contrast, side, are a hundred times quicker than CPUs for Neural networks, however, they are prohibitively expensive compared to CPUs. Using the systolic array structure, TPU performs well on neural networks with high batch sizes but in GAN the shift between CPU and TPU is huge so it does not perform well.

关键词： Generative Adversarial Network High-Performance computing Graphic Processing Unit Tensor Processing Unit

来源：评论

学校读者我要写书评

暂无评论

A distributed Ensemble of Diverse Deep Learning Models for Predicting COVID-19 Cases

A Distributed Ensemble of Diverse Deep Learning Models for P...

引用

Eurocon

作者： Mohammad Mehedi Hassan Mabrook S. AlRakhami Ahmed Zohier Elhendi Salman A. AlQahtani Information Systems Department CCIS King Saud University Riadh Saudi Arabia Science Technology and Innovation Dept. King Saud University Riyadh Saudi Arabia Computer Engineering Department CCIS King Saud University Riadh Saudi Arabia

The outbreak of the COVID-19 pandemic has resulted in a significant impact on global health and economy. Forecasting the spread of COVID-19 cases is essential for policymakers to make informed decisions and allocate resources accordingly. Deep learning has shown promising results in predicting the spread of infectious diseases, including COVID-19. However, selecting the most appropriate model for a given forecasting task can be challenging, as no single model is guaranteed to perform well for all scenarios. In this paper, we propose a distributed ensemble approach that combines the predictions of different deep learning models, including Autoregressive Integrated Moving Average (Auto-ARIMA), Convolutional Neural networks (CNN) and Long Short-Term Memory (LSTM), to improve the accuracy of COVID-19 case forecasts. Specifically, we train each deep learning model on a subset of the data and aggregate their predictions to obtain an ensemble forecast. To achieve a distributed ensemble, we use an edge computing framework to train the deep learning models on multiple nodes in a parallel and distributed manner. We evaluate our approach on publicly available COVID-19 datasets, and our results show that the proposed ensemble model outperforms the individual models in terms of forecasting accuracy. Furthermore, our distributed ensemble approach significantly reduces the training and prediction time compared to training each model separately.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Partition Pruning: parallelization-Aware Pruning for Dense Neural networks 28

Partition Pruning: Parallelization-Aware Pruning for Dense N...

引用

28th Euromicro international conference on parallel, distributed and Network-Based Processing (PDP)

作者： Shahhosseini, Sina Albaqsami, Ahmad Jasemi, Masoomeh Bagherzadeh, Nader Univ Calif Irvine Elect Engn & Comp Sci Dept Irvine CA 92697 USA Sharif Univ Technol Dept Comp Engn Tehran Iran

ISBN: (纸本)9781728165820

As recent neural networks are being improved to be more accurate, their model's size is exponentially growing. Thus, a huge number of parameters requires to be loaded and stored from/in memory hierarchy and computed in processors to perform training or inference phase of neural network processing. Increasing the number of parameters causes a big challenge for real-time deployment since the memory bandwidth improvement's trend cannot keep up with models' complexity growing trend. Although some operations in neural networks processing are computational intensive such as convolutional layer computing, computing dense layers face with memory bandwidth bottleneck. To address the issue, the paper has proposed Partition Pruning for dense layers to reduce the required parameters while taking into consideration parallelization. We evaluated the performance and energy consumption of parallel inference of partitioned models, which showed a 7.72x speedup of performance and a 2.73x reduction in the energy used for computing pruned fully connected layers in TinyVGG16 model in comparison to running the unpruned model on a single accelerator. Besides, our method showed a limited reduction in accuracy while partitioning fully connected layers.

关键词： parallelization Deep Neural Network Pruning Partitioning Hardware Accelerator

来源：评论

学校读者我要写书评

暂无评论

Proactive Failure Recovery for Stateful NFV 26

Proactive Failure Recovery for Stateful NFV

引用

26th IEEE international conference on parallel and distributed Systems (IEEE ICPADS)

作者： Huang, Zhenyi Huang, Huawei Sun Yat Sen Univ Sch Data & Comp Sci Guangzhou Peoples R China

ISBN: (纸本)9781728190747

Network Function Virtualization (NFV) technology is viewed as a significant component of both the fifth-generation (5G) communication networks and edge computing. In this paper, through reviewing the state-of-the-art work on applying NFV to edge computing, we identify that an urgent research challenge is to provide the proactive failure recovery mechanism for the stateful NFV. To realize such proactive failure recovery, we propose a prediction-based algorithm for redeploying the stateful NFV instances in real-time when network failures occur. The proposed algorithm is based on relax and rounding technique. The theoretical performance guarantee is also analyzed rigorously. Simulation results show that the proposed failure recovery algorithm outperforms the reactive-manner baselines significantly in terms of redeployment latency.

关键词： NFV Failure Recovery Edge computing

来源：评论

学校读者我要写书评

暂无评论

Resource Allocation Strategy for Virtualized Wireless Sensor networks Based on Actor-Critic

Resource Allocation Strategy for Virtualized Wireless Sensor...

引用

IEEE international Symposium on parallel and distributed Processing with Applications and IEEE international conference on Ubiquitous computing and Communications (ISPA/IUCC)

作者： Changjin Li Guangxu Zhou Yue Guo Junfeng Hou Hongjun Zhang Song Li Xuchang Branch of Henan Provincial Tobacco Company Xuchang Henan China

Aiming at the problem of effective utilization of virtualized WSN resources, a reinforcement learning-based resource allocation strategy for virtualized WSNs is proposed from the perspective of service requests of virtualized WSNs. A virtual sensor network request deployment utility maximization model is established based on the quality requirements of service sensing information and resource capacity constraints. Considering the randomness of service requests and the dynamic changes in the network environment, the resource optimization problem is transformed into MDP and solved using Actor-Critic reinforcement learning methods, and the best resource allocation strategy is obtained. Finally, a large number of simulation experiments are carried out to verify the effectiveness of the proposed method.

关键词： Wireless sensor networks Soft sensors Computational modeling Reinforcement learning Ubiquitous computing Dynamic scheduling Resource management

来源：评论

学校读者我要写书评

暂无评论

MULTI-GPU TRAINING and parallel CPU computing for the MACHINE LEARNING EXPERIMENTS USING ARIADNE LIBRARY 9

MULTI-GPU TRAINING and PARALLEL CPU COMPUTING for the MACHIN...

引用

9th international conference "distributed computing and Grid Technologies in Science and Education", GRID 2021

作者： Goncharov, Pavel Nikolskaia, Anastasiia Ososkov, Gennady Rezvaya, Ekaterina Rusov, Daniil Shchavelev, Egor Joint Institute for Nuclear Research 6 Joliot-Curie Street Dubna Moscow region 141980 Russia Saint Petersburg State University 7-9 Universitetskaya emb. Saint Petersburg199034 Russia

Modern machine learning (ML) tasks and neural network (NN) architectures require huge amounts of GPU computational facilities and demand high CPU parallelization for data preprocessing. At the same time, the Ariadne library, which aims to solve complex high-energy physics tracking tasks with the help of deep neural networks, lacks multi-GPU training and efficient parallel data preprocessing on the CPU. In our work, we present our approach for the Multi-GPU training in the Ariadne library. We will present efficient data-caching, parallel CPU data preprocessing, generic ML experiment setup for prototyping, training, and inference deep neural network models. Results in terms of speed-up and performance for the existing neural network approaches are presented with the help of GOVORUN computing resources. Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 international (CC BY 4.0).

关键词： Graphics processing unit

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：