检索结果-内蒙古大学图书馆

2023 World Wide Web Conference, WWW 2023

作者： Yuan, Liang He, Qiang Chen, Feifei Dou, Ruihan Jin, Hai Yang, Yun School of Computer Science and Technology Huazhong University of Science and Technology China Department of Computing Technologies Swinburne University of Technology Australia School of Information Technology Deakin University Australia Faculty of Mathematics University of Waterloo Canada National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Huazhong University of Science and Technology Wuhan430074 China

ISBN: (纸本)9781450394161

Powered by the massive data generated by the blossom of mobile and Web-of-Things (WoT) devices, Deep Neural Networks (DNNs) have developed both in accuracy and size in recent years. Conventional cloud-based DNN training incurs rapidly-increasing data and model transmission overheads as well as privacy issues. Mobile edge computing (MEC) provides a promising solution by facilitating DNN model training on edge servers at the network edge. However, edge servers often suffer from constrained resources and need to collaborate on DNN training. Unfortunately, managed by different telecoms, edge servers cannot properly collaborate with each other without incentives and trust. In this paper, we introduce PipeEdge, a scheme that promotes collaborative edge training between edge servers by introducing incentives and trust based on blockchain. Under the PipeEdge scheme, edge servers can hire trustworthy workers for pipelined DNN training tasks based on model parallelism. We implement PipeEdge and evaluate it comprehensively with four different DNN models. The results show that it outperforms state-of-the-art schemes by up to 173.98% with negligible overheads. © 2023 ACM.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Dependability Analysis for Fault-Tolerant Computer systems Using Dynamic Fault Graphs

引用

China Communications 2014年第9期11卷 16-30页

作者： ZHAO Feng JIN Hai ZOU Deqing QIN Pan Services Computing Technology and System Lab & Cluster and Grid Computing Lab School of Computer Scienceand Technology Huazhong University of Science and Technology Wuhan 430074 Hubei Province P. R. China

Dependability analysis is an important step in designing and analyzing safety computer systems and protection *** multi-processor and virtual machine increases the system faults' complexity,diversity and dynamic,in particular for software-induced failures,with an impact on the overall ***,it is very different for safety system to operate successfully at any active phase,since there is a huge difference in failure rate between hardware-induced and softwareinduced *** handle these difficulties and achieve accurate dependability evaluation,consistently reflecting the construct it measures,a new formalism derived from dynamic fault graphs(DFG) is developed in this *** exploits the concept of system event as fault state sequences to represent dynamic behaviors,which allows us to execute probabilistic measures at each timestamp when change *** approach automatically combines the reliability analysis with the system *** this paper,we describe how to use the proposed methodology drives to the overall system dependability analysis through the phases of modeling,structural discovery and probability analysis,which is also discussed using an example of a virtual computing system.

关键词： fault-tolerant system depend-ability analysis dynamic fault-graph structural link probability forecast

来源：评论

学校读者我要写书评

暂无评论

MR-COF: A genetic MapReduce configuration optimization framework 15th

MR-COF: A genetic MapReduce configuration optimization frame...

引用

15th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2015

作者： Liu, Chao Zeng, Deze Yao, Hong Hu, Chengyu Yan, Xuesong Fan, Yuanyuan Hubei Key Laboratory of Intelligent Geo-Information Processing China University of Geosciences Wuhan430074 China China Services Computing Technology and System Lab and Cluster and Grid Computing Lab Huazhong University of Science and Technology Wuhan430074 China

ISBN: (纸本)9783319271392

Hadoop/MapReduce has emerged as a de facto programming framework to explore cloud-computing resources. Hadoop has many configuration parameters, some of which are crucial to the performance of MapReduce jobs. In practice, these parameters are usually set to default or inappropriate values. This severely limits system performance (e.g., execution time). Therefore, it is essential but also challenging to investigate how to automatically tune these parameters to optimize MapReduce job performance. In this paper, we propose an automatic MapReduce configuration optimization framework named as MR-COF. By monitoring and analyzing the runtime behavior, the framework adopts a cost-based performance prediction model that predicts the MapReduce job performance. In addition, we design a genetic search algorithm which iteratively tunes parameters in order to find out the best one. Testbed-based experimental results show that the average MapReduce job performance is increased by 35% with MR-COF compared to the default configuration. © Springer International Publishing Switzerland 2015.

关键词： MapReduce

来源：评论

学校读者我要写书评

暂无评论

xGCN: An Extreme Graph Convolutional Network for Large-scale Social Link Prediction 23

xGCN: An Extreme Graph Convolutional Network for Large-scale...

引用

2023 World Wide Web Conference, WWW 2023

作者： Song, Xiran Lian, Jianxun Huang, Hong Luo, Zihan Zhou, Wei Lin, Xue Wu, Mingqi Li, Chaozhuo Xie, Xing Jin, Hai Huazhong University of Science and Technology Wuhan China Microsoft Research Asia Beijing China Microsoft Gaming Redmond United States National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology China

ISBN: (纸本)9781450394161

Graph neural networks (GNNs) have seen widespread usage across multiple real-world applications, yet in transductive learning, they still face challenges in accuracy, efficiency, and scalability, due to the extensive number of trainable parameters in the embedding table and the paradigm of stacking neighborhood aggregations. This paper presents a novel model called xGCN for large-scale network embedding, which is a practical solution for link predictions. xGCN addresses these issues by encoding graph-structure data in an extreme convolutional manner, and has the potential to push the performance of network embedding-based link predictions to a new record. Specifically, instead of assigning each node with a directly learnable embedding vector, xGCN regards node embeddings as static features. It uses a propagation operation to smooth node embeddings and relies on a Refinement neural Network (RefNet) to transform the coarse embeddings derived from the unsupervised propagation into new ones that optimize a training objective. The output of RefNet, which are well-refined embeddings, will replace the original node embeddings. This process is repeated iteratively until the model converges to a satisfying status. Experiments on three social network datasets with link prediction tasks show that xGCN not only achieves the best accuracy compared with a series of competitive baselines but also is highly efficient and scalable. © 2023 ACM.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Autonomous Agent Based Intrusion Detection in Virtual computing Environment

Autonomous Agent Based Intrusion Detection in Virtual Comput...

引用

2010 IEEE International Conference Conferenhce on Wireless Communications,Networking and Information Security(2010 IEEE 无线通信、网络技术与信息安全国际会议 WCNIS)

作者： Feng Zhao Alex K. Ohoussou Deqing Zou Hai Jin Guofu Xiang Ge Cheng Cluster and Grid Computing Lab Services Computing Technology and System Lab Huazhong University of S Cluster and Grid Computing Lab Services Computing Technology and System Lab Huazhong University of S

One of the motivations for virtualization technology is the desire to develop new services to enhance system security without trusting both the applications and the operating systems. An intrusion detection system is an example of such service that can help to isolate users from malicious attacks. In this paper, we propose hybrid-based intrusion detection architecture in virtual computing environment to detect and isolate harmful behaviors by real-time monitoring and alarming. In contrast to monolithic intrusion detection system, we introduce autonomous agents, acting independently of each other, to monitor the system. The agents are deployed in virtual machines to analyze actions occurring on the network and inside the hosts to determine whether they are potential security violations or not. Our architecture is implemented based on Xen, and the detection management center is deployed in a secure virtual machine.

关键词： intrusion detection autonomous agent virtual computing environment

来源：评论

学校读者我要写书评

暂无评论

Optimization of asynchronous graph processing on GPU with hybrid coloring model 2015

Optimization of asynchronous graph processing on GPU with hy...

引用

20th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2015

作者： Shi, Xuanhua Liang, Junling Di, Sheng He, Bingsheng Jin, Hai Lu, Lu Wang, Zhixiang Luo, Xuan Zhong, Jianlong Services Computing Technology and System Lab. Cluster and Grid Computing Lab. School of Computer Science and Technology Huazhong University of Science and Technology China Argonne National Laboratory United States School of Computer Engineering Nanyang Technological University Singapore Singapore

ISBN: (纸本)9781450332057

Modern GPUs have been widely used to accelerate the graph processing for complicated computational problems regarding graph theory. Many parallel graph algorithms adopt the asynchronous computing model to accelerate the iterative convergence. Unfortunately, the consistent asynchronous computing requires locking or the atomic operations, leading to significant penalties/overheads when implemented on GPUs. To this end, coloring algorithm is adopted to separate the vertices with potential updating conflicts, guaranteeing the consistency/correctness of the parallel processing. We propose a light-weight asynchronous processing framework called Frog with a hybrid coloring model. We find that majority of vertices (about 80%) are colored with only a few colors, such that they can be read and updated in a very high degree of parallelism without violating the sequential consistency. Accordingly, our solution will separate the processing of the vertices based on the distribution of colors.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

A measurement-based study on user management in private BitTorrent communities

A measurement-based study on user management in private BitT...

引用

作者： Jiang, Honglei Jin, Hai Guo, Song Liao, Xiaofei Services Computing Technology and System Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China School of Computer Science and Engineering University of Aizu Aizu-Wakamatsu Japan

Beyond the traditional BitTorrent, a new genre of peer-to-peer communications protocol for worldwide file sharing is rapidly evolving towards private BitTorrent (PT). In recent years, a proliferation of PT communities have emerged. To enhance the user experience, account-based share-ratio enforcement (SRE) has been developed and widely adopted. Whereas existing studies mainly take SRE as an incentive, we discover that it also plays a critical role in selecting and filtering users. In addition to SRE, a rich set of user management rules, such as registration management, banning policies, and user caste system, are also studied in this paper. This includes to explore their effects on user behavior, download performance, content availability, and system scalability. The measurement results presented in this paper are based on large-scale experiments conducted over six representative PT sites for over a year. We find that the stricter registration will lead to fewer new users, resulting in a scalability problem, which is critical for the PT communities because the download performance and content availability depend on not only the contribution of users but also the population of the community. Our measurement and analysis pose a direction for the design of new incentive mechanisms that take the difficulty of enrollment into the *** © 2012 John Wiley & Sons, Ltd. Copyright © 2012 John Wiley & Sons, Ltd.

关键词： Behavioral research

来源：评论

学校读者我要写书评

暂无评论

Virtual machine management based on agent service

Virtual machine management based on agent service

引用

11th International Conference on Parallel and Distributed computing, Applications and Technologies, PDCAT 2010

作者： Wu, Song Deng, Li Jin, Hai Shi, Xuanhua Zhao, Yankun Gao, Wei Zhang, Jianyin Peng, Jin Services Computing Technology and System Lab. School of Computer Science and Technology Huazhong University of Science and Technology Wuhan 430074 China China Mobile Research Institute Beijing 100053 China

ISBN: (纸本)9780769542874

With the popularity of virtualization, the problem that how to manage hundreds even thousands of virtual machines running on multiple physical computing nodes becomes important. Current virtual machine management systems only can obtain basic information of virtual machines and execute simple operations on them, such as start, reboot and shutdown. In this paper, we design a virtual machine management approach based on agent service. Agent service can provide detail running status information inside virtual machines. It also has been a bridge for host machines and virtual machines to interact with each other. Agent service is designed to automatically start when virtual machine boots up. By agent service we can get real-time information about virtual machines. We evaluate monitoring overhead and the performance of batch operations when using agent. The experimental results show that agent mechanism outperforms methods using Libvirt or VMware tools. © 2010 IEEE.

关键词： Virtual machine

来源：评论

学校读者我要写书评

暂无评论

Minimum Spanning Tree Clustering Based on Density Filtering 7th

Minimum Spanning Tree Clustering Based on Density Filtering

引用

7th CCF Academic Conference on BigData, CCF BigData 2019

作者： Wang, Ke Xie, Xia Sun, Jiayu Cao, Wenzhi National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Hunan University of Technology and Business Changsha410205 China

ISBN: (纸本)9789811518980

Clustering analysis is an important method in data mining. In order to recognize clusters with arbitrary shapes as well as clusters with different density, we propose a new clustering approach: minimum spanning tree clustering based on density filtering. It masks the low-density points in the density filtering step, which reduces the interference of noise and makes the gap between clusters clearer. It uses relative values of adjacent distances to find mutations of density and changes between clusters to divide data sets. It is tested on multiple synthetic data sets and real-world data sets, the results of which show that the algorithm is able to detect clusters with arbitrary shape and it is insensitive to the imbalance of density between clusters. It has achieved great results on multiple data sets. © Springer Nature Singapore Pte Ltd 2019.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

CUDAGA: A portable parallel programming model for GPU cluster 2nd

CUDAGA: A portable parallel programming model for GPU cluste...

引用

2nd International Conference on Cloud computing and Big Data, CloudCom-Asia 2015

作者： Chen, Yong Jin, Hai Xu, Dechao Zheng, Ran Liu, Haocheng Zeng, Jingxiang Services Computing Technology and System Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China China Electric Power Research Institute Beijing100192 China

ISBN: (纸本)9783319284293

GPU cluster is important for high performance computing with its high performance/cost ratio. However, it is still very hard for application developers to write parallel codes on GPU. MPI is mostly used for parallel programming, and data locality and communication must be specified explicitly by developers. Moreover, data transmission between CPU and GPU must also be processed with CUDA codes. CUDAGA, a new parallel programming model for GPU cluster with CUDA, is presented to provide portable interfaces for communication on GPUs. GA (Global Arrays), a portable shared-memory programming model for distributed memory computers, is the base to facilitate parallel programming and maintain transparent global arrays on GPUs. Experiments show that CUDAGA can decrease parallel programming difficulties, but ensures better performance for some specific applications. © Springer International Publishing Switzerland 2015.

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：