检索结果-内蒙古大学图书馆

International Conference on Networking, Architecture, and Storage (NAS)

作者： Zhaoning Zhang Kui Wu Huiba Li Jinghua Feng Yuxing Peng Xicheng Lu National Key Laboratory for Parallel and Distributed Processing National University of Defense Technology China Computer Science Department University of Victoria B.C. Canada National Super Computing Center Tianjin China

Virtualization is the foundation for cloud computing, and the virtualization can not be achieved without software defined, elastic, flexible and scalable virtual layers. Unfortunately, if multiple virtual storage devices are chained together, the system may be subject to severe performance degradation. While the read-ahead (RA) mechanism in storage devices plays a very important role to improve I/O performance, RA may not be effective as expected for multiple virtualization layers, since it is originally designed for one layer only. When I/O requests are passed through a long I/O path, they may trigger a chain reaction and lead to unnecessary data transmission and thus bandwidth waste. In this paper, we study the dynamic behavior of RA through multiple I/O layers and demonstrate that if controlled well, RA can greatly accelerate I/O speed. We present RAFlow, a RA control mechanism, to effectively improve I/O performance by strategically expanding RA window at each layer. Our real-world experiments show that it can achieve 20% to 50% performance improvement in I/O paths with up to 8 virtualized storage devices.

关键词： Radio access networks Acceleration distributed databases Performance evaluation Synchronization Benchmark testing Degradation

来源：评论

学校读者我要写书评

暂无评论

Accuracy evaluation of deep belief networks with fixed-point arithmetic

引用

Computer Modelling and New Technologies 2014年第6期18卷 7-14页

作者： Jiang, Jingfei Hu, Rongdong Mikel, Lujan Dou, Yong Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology ChangSha Hunan China University of Manchester Manchester United Kingdom

Deep Belief Networks (DBNs) are state-of-art Machine Learning techniques and one of the most important unsupervised learning algorithms. Training DBNs is computationally intensive which naturally leads to investigate FPGA acceleration. Fixed-point arithmetic can be used when implementing DBNs in FPGAs to reduce execution time, but it is not clear the implications for accuracy. Previous studies have focused only on accelerators using some fixed bit-widths. A contribution of this paper is to demonstrate the bit-width effect on various configurations of DBNs in a comprehensive way by experimental evaluation. Explicit performance changing points are found using various bit-widths. The impact of sigmoid function approximation, required part of DBNs, is evaluated. A solution of mixed bit-widths DBN is proposed, fitting the bit-widths of FPGA primitives and gaining similar performance to the software implementation. Our results provide a guide to inform the design choices on bit-widths when implementing DBNs in FPGAs documenting clearly the trade-off in accuracy.

关键词： Field programmable gate arrays (FPGA)

来源：评论

学校读者我要写书评

暂无评论

An abstract domain to infer octagonal constraints with absolute value

Lecture Notes in Computer Science (including subseries Lectu...

引用

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 2014年 8723卷 101-117页

作者： Chen, Liqian Liu, Jiangchao Mine, Antoine Kapur, Deepak Wang, Ji National Laboratory for Parallel and Distributed Processing National University of Defense Technology Changsha China École Normale Supérieure Paris France CNRS France University of New Mexico NM United States

The octagon abstract domain, devoted to discovering octagonal constraints (also called Unit Two Variable Per Inequality or UTVPI constraints) of a program, is one of the most commonly used numerical abstractions in practice, due to its quadratic memory complexity and cubic time complexity. However, the octagon domain itself is restricted to express convex sets and has limitations in handling non-convex properties which are sometimes required for proving some numerical properties in a program. In this paper, we intend to extend the octagon abstract domain with absolute value, to infer certain non-convex properties by exploiting the absolute value function. More precisely, the new domain can infer relations of the form {±X ± Y ≤ c, ±X ± |Y| ≤ d,±|X|±|Y| ≤ e}. We provide algorithms for domain operations such that the new domain still enjoys the same asymptotic complexity as the octagon domain. Moreover, we present an approach to support strict inequalities over rational or real-valued variables in this domain, which also fits for the octagon domain. Experimental results of our prototype are encouraging;The new domain is scalable and able to find non-convex invariants of interest in practice but without too much overhead (compared with that using octagons). © Springer International Publishing Switzerland 2014.

关键词： Set theory

来源：评论

学校读者我要写书评

暂无评论

Power and performance analysis of the Graph 500 benchmark on the Single-chip Cloud Computer

Power and performance analysis of the Graph 500 benchmark on...

引用

International Conference on Cloud Computing and Internet of Things (CCIOT)

作者： Zhiquan Lai King Tin Lam Cho-Li Wang Jinshu Su College of Computer National University of Defense Technology Changsha China Department of Computer Science The University of Hong Kong Hong Kong China National Key Laboratory of Parallel and Distributed Processing (PDL) Changsha China

ISBN: (纸本)9781479947638

The concerns of data-intensiveness and energy awareness are actively reshaping the design of high-performance computing (HPC) systems nowadays. The Graph500 is a widely adopted benchmark for evaluating the performance of computing systems for data-intensive workloads. In this paper, we introduce a data-parallel implementation of Graph500 on the Intel Single-chip Cloud Computer (SCC). The SCC features a non-coherent many-core architecture and multi-domain on-chip DVFS support for dynamic power management. With our custom-made shared virtual memory programming library, memory sharing among threads is done efficiently via the shared physical memory (SPM) while the library has taken care of the coherence. We conduct an in-depth study on the power and performance characteristics of the Graph500 workloads running on this system with varying system scales and power states. Our experimental results are insightful for the design of energy-efficient many-core systems for data-intensive applications.

关键词： Kernel Benchmark testing Computer architecture Programming Power measurement Computational modeling Runtime

来源：评论

学校读者我要写书评

暂无评论

Automatic Detecting Performance Bugs in Cloud Computing Systems via Learning Latency Specification Model

Automatic Detecting Performance Bugs in Cloud Computing Syst...

引用

2014 IEEE 8th International Symposium on Service Oriented System Engineering

作者： Haibo Mi Huaimin Wang Zhenbang Chen Yangfan Zhou Science and Technology on Parallel & Distributed Processing Laboratory National Univ. of Defense Technology Shenzhen Research Institute The Chinese Univ. of Hong Kong

Performance bugs that don't cause fail-stop errors but degradation of system performance have been one of the most fundamental issues in the production platform. How to effectively online detect bugs becomes more and more urgent for engineers. Performance bugs usually manifest themselves as the anomalous call structures of request traces or anomalous latencies of invoked methods. In this paper, we propose an automatic performance bug online detecting approach, CloudDoc. CloudDoc maintains a performance model mined from execution traces that are collected in the normal period. The performance model captures the characteristics of call structures of request traces together with corresponding latencies. With the performance model, CloudDoc periodically detects whether performance bugs occur or not. All suspicious call structures or latency-abnormal invoked methods are presented to engineers. We report two case studies to demonstrate the effectiveness of CloudDoc in helping engineers identify performance bugs.

关键词： Computer bugs Computational modeling Cloud computing Electronic mail Merging System performance Servers

来源：评论

学校读者我要写书评

暂无评论

Who Should Review this Pull-Request: Reviewer Recommendation to Expedite Crowd Collaboration

Who Should Review this Pull-Request: Reviewer Recommendation...

引用

Asia-Pacific Conference on Software Engineering

作者： Yue Yu Huaimin Wang Gang Yin Charles X. Ling National Laboratory for Parallel and Distributed Processing College of Computer National University of Defense Technology Changsha China Department of Computer Science The University of Western Ontario London Ontario Canada

Github facilitates the pull-request mechanism as an outstanding social coding paradigm by integrating with social media. The review process of pull-requests is a typical crowd sourcing job which needs to solicit opinions of the community. Recommending appropriate reviewers can reduce the time between the submission of a pull-request and the actual review of it. In this paper, we firstly extend the traditional Machine Learning (ML) based approach of bug triaging to reviewer recommendation. Furthermore, we analyze social relations between contributors and reviewers, and propose a novel approach to recommend highly relevant reviewers by mining comment networks (CN) of given projects. Finally, we demonstrate the effectiveness of these two approaches with quantitative evaluations. The results show that CN-based approach achieves a significant improvement over the ML-based approach, and on average it reaches a precision of 78% and 67% for top-1 and top-2 recommendation respectively, and a recall of 77% for top-10 recommendation.

关键词： Communities Rails Training Software Encoding Mathematical model Social network services

来源：评论

学校读者我要写书评

暂无评论

Enhancing the Security of parallel Programs via Reducing Scheduling Space

Enhancing the Security of Parallel Programs via Reducing Sch...

引用

IEEE International Symposium on Dependable, Autonomic and Secure Computing (DASC)

作者： Xu Zhou Gen Li Kai Lu Shuangxi Wang School of Computer National University of Defense Technology Changsha PR China Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology Changsha PR China

parallel programs face a new security problem - concurrency vulnerability, which is caused by a special thread scheduling instead of inputs. In this paper, we propose to automatically fix concurrency vulnerabilities by reducing thread scheduling space. Our method is based on two observations. First, most concurrency vulnerabilities are caused by atomicity violation errors. Second, reducing thread scheduling space does not harm the correctness of the original program. We designed a prototype runtime system shield using deterministic multithreading techniques. Shield is designed to transparently run parallel programs and schedule threads in large instruction blocks to prevent atomicity violation at best effort. In case some concurrency vulnerabilities cannot be fixed by shield's scheduling reducing scheme, we also provide a remedy strategy by integrating shield with record&replay function, so that it can help programmers to analyze attacker's behavior for manually fixing.

关键词： Concurrent computing Instruction sets Scheduling Runtime Computer bugs Security Processor scheduling

来源：评论

学校读者我要写书评

暂无评论

A cooperative CSMA/CA MAC for channel-hopping rendezvous based cognitive radio networks

A cooperative CSMA/CA MAC for channel-hopping rendezvous bas...

引用

International Conference on Communications and Networking in China (CHINACOM)

作者： Quan Liu Gang Hu Xiaodong Wang Xingming Zhou Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology Changsha Hunan China Department of Network Engineering National University of Defense Technology Changsha Hunan China

ISBN: (纸本)9781479959716

Rendezvous problem is known as the fundamental issue in MAC design for cognitive radio networks (CRNs). With the concept of blind rendezvous, numerous channel-hopping sequence (CHS) based rendezvous schemes have been proposed to solve this problem. However, little attention is paid to the design of a CSMA/CA MAC based on these rendezvous schemes. In this paper, we propose a cooperative CSMA/CA MAC (named CoCH-CSMA/CA MAC) which tailors 802.11 distributed coordination function (DCF) to the slotted operation manner of existing CHS based rendezvous schemes. More importantly, the rendezvous de-synchronization problem is identified in MAC design based on these rendezvous schemes. To alleviate its impact on MAC performance, a cooperative control feedback scheme employing correlation-based signal detection is proposed to help secondary users avoid backoff misbehavior and improve networking performance (i.e., packet delivery delay and network throughput). Extensive simulations are conducted to prove the effectiveness of our MAC design.

关键词： Delays Multiaccess communication Correlation Radiation detectors IEEE 802.11 Standards Receivers

来源：评论

学校读者我要写书评

暂无评论

Assessing the Impact of Cascading Failures on the Interdomain Routing System of the Internet

引用

New Generation Computing 2014年第3期32卷 237-255页

作者： Liu, Yujing Peng, Wei Su, Jinshu Wang, Zhilin College of Computer National University of Defense Technology Hunan Changsha China College of Computer Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology Hunan Changsha China Education Department National University of Defense Technology Hunan Changsha China

The Internet is a typical complex network, whose traffic load is controlled by the inter-domain routing system. Due to the co-location of data plane and control plane of Border Gateway Protocol, the survivability of inter-domain routing system is sensitive to severe congestion. Therefore, an initial outage may lead to a cascade of failures in the Internet. But the cascading failures on links are able to be automatically restored when the congestion is mitigated. In this paper, we propose a model - CAFEIN for characterizing this special process. Based on CAFEIN, we assess the difference of impact under intentional attacks and random breakdowns; identify the worst affected part of the Internet; and study the propagation of cascading failures. Through simulations, we find that the cascading failures bring a great deal of added burden to the routing system. However, the cascading effect is amplified globally when the relative capacity of links is very low. Moreover, the difference of impact between intentional attack and random breakdown is not as prominent as previous research due to the unique automatic-restoration process.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Effective hierarchical routing algorithm for large-scale wireless mobile networks

引用

Journal of Networks 2014年第2期9卷 344-352页

作者： Yan, Guofeng Peng, Yuxing Hu, Ying Liu, Junyi School of Computer and Communication Hunan Institute of Engineering Hunan 411104 China Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology Changsha 410073 China School of Computer National University of Defense Technology Changsha 410073 China

The growing interest in wireless mobile network techniques has resulted in many routing protocol proposals. The unpredictable motion and the unreliable behavior of mobile nodes is one of the key issues in wireless mobile network. Virtual mobile node (VMN) consists of robust virtual nodes that are both predictable and reliable. Based on VMN, in this paper, we present a hierarchical routing algorithm, i.e., EHRA-WAVE, for large-scale wireless mobile networks. By using mobile WAVE technology, a routing path can be found rapidly between VMNs without accurate topology information. We introduce the routing algorithm and the implementation issues of the proposed EHRA-WAVE routing algorithm. Finally, we evaluate the performance of EHRA-WAVE through experiments, and compare the performance on VMN failure and message delivery ratio using hierarchical and non-hierarchical routing methods. However, due to the large amounts WAVE flooding, EHRA-WAVE results in too large load which would impede the application of the EHRA-WAVE algorithm. Therefore, the further routing protocol focuses on minimizing the number of WAVE using hierarchical structures in large-scale wireless mobile networks. © 2014 Academy Publisher.

关键词： Quality of service

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：