检索结果-内蒙古大学图书馆

SORTING AND ROUTING ON OTIS-MESH OF TREES

parallel PROCESSING LETTERS 2010年第2期20卷 145-154页

作者： Lucas, Keny T. Jana, Prasanta K. Xavier Inst Social Serv Dept Informat Management Ranchi 834001 Bihar India Mines Univ Indian Sch Dept Comp Sci & Engn Dhanbad 826004 Bihar India

OTIS (optical transpose interconnection system) is a popular model of optoelectronic parallel computers that has gained enormous attention in the recent years. Several parallel algorithms have been published for many fundamental problems on this architecture. In this paper, first we propose a parallel algorithm for sparse enumeration sort on OTISMesh of Trees (OTIS-MOT). For N (= n(2)) data elements, our sorting algorithm requires 4 log N ($) electronic moves + 3 OTIS moves. We next present a shortest path routing algorithm that runs also in logarithmic time.

关键词： Optical transpose interconnection system (OTIS) interconnection network parallel algorithm enumeration sort shortest path routing

来源：评论

学校读者我要写书评

暂无评论

Observations on optimal parallelizations of two-list algorithm

引用

parallel COMPUTING 2010年第1期36卷 65-67页

作者： Alonso Sanches, Carlos Alberto Soma, Nei Yoshihiro Yanasse, Horacio Hideki CTA ITA IEC Inst Tecnol Aeronaut BR-12228900 Sao Jose Dos Campos SP Brazil INPE LAC Inst Nacl Pesquisas Espaciais BR-12227010 Sao Jose Dos Campos SP Brazil

For more than three decades, the very well known and famous two-list Horowitz and Sahni algorithm [3] remains the serial upper-bound for the 0-1 Knapsack problem with n items (KP01) in a time bounded by O(2(n/2)). Recently, Chedid [2] Suggested an optimal parallelization for that algorithm to a KP01 variation - the subset-sum problem - in a PRAM CREW with p = 2(n/8) processors. It is presented here that, in addition to be incomplete, the Chedid result is a particular case given by Sanches et al. [6]. (c) 2009 Elsevier B.V. All rights reserved.

关键词： Knapsack problem Subset-sum problem parallel algorithm Two-list algorithm

来源：评论

学校读者我要写书评

暂无评论

INTELLIGENT TECHNOLOGIES OF HIGH-PERFORMANCE COMPUTING

引用

CYBERNETICS AND SYSTEMS ANALYSIS 2010年第5期46卷 833-844页

作者： Sergienko, I. V. Molchanov, I. N. Khimich, A. N. Natl Acad Sci Ukraine VM Glushkov Cybernet Inst Kiev Ukraine

This article analyzes mathematical and technological problems that arise in performing computational experiments on modern high-performance computers (supercomputers). As a means of overcoming the difficulties associated with the analysis and solution of computer model problems under conditions of approximate initial data on computers with parallel architectures, intelligent technologies are proposed that are based on intelligent software supported by architectural decisions of an intelligent computer and predictive system software.

关键词： approximate initial data reliability of a solution parallel algorithm intelligent software intelligent MIMD computer

来源：评论

学校读者我要写书评

暂无评论

parallel algorithms for Solving Markov Decision Process

引用

9th International Conference on algorithms and Architectures for parallel Processing

作者： Zhang, Qi Sun, Guangzhong Xu, Yinlong Univ Sci & Technol China Dept Comp Sci Hefei 230027 Anhui Peoples R China

ISBN: (纸本)9783642030949

Markov decision process (MDP) provides the foundations for a number of problems, such as artificial intelligence Studying, automated planning and reinforcement learning. MDP can be solved efficiently in theory. However, for large scenarios, more investigations are needed to reveal practical algorithms. algorithms for solving MDP have a natural concurrency. In this paper, we present parallel algorithms based on dynamic programming Meanwhile, the cost of computation and communication complexity of this method is analyzed. Moreover, experimental results demonstrate excellent speedups and scalability.

关键词： Markov Decision Process (MDP) parallel algorithm

来源：评论

学校读者我要写书评

暂无评论

parallel algorithms for Constructing Fellow Automata of Regular Expressions

Parallel Algorithms for Constructing Fellow Automata of Regu...

引用

1st International Workshop on Database Technology and Applications

作者： Sun Yuqiang Yang Ruimin Gu Yuwan You Jing Jiangsu Polytech Univ Sch Informat Sci & Engn Changzhou 213164 South Korea Henan Normal Univ Coll Comp & Informat Technol Xinxiang 453007 Peoples R China

ISBN: (纸本)9780769536040

A parallel algorithm for translating regular expression into its Follow automata is proposed in the paper. Firstly, we construct the Thompson automata of a regular expression. Then, the Glushkov automata are achieved by removing the xi path and the equivalent states which have equivalent relations are merged into one. So we can get a smaller finite automata, named Follow automata. Finally the parallel processing of algorithm is described in detail with an example.

关键词： NFA state regular expressions parallel algorithm Follow automatat

来源：评论

学校读者我要写书评

暂无评论

Efficient parallel adaptive array beamforming algorithm

引用

Journal of Systems Engineering and Electronics 2009年第6期20卷 1221-1226页

作者： Huang Fei Sheng Weixing Ma Xiaofeng Dept. of Communication Engineering Nanjing Univ. of Science and Technology Nanjing 210094 P. R. China

For a large-scale adaptive array, the heavy computational load and the high-rate data transmission are two challenges in the implementation of an adaptive digital beamforming system. An efficient parallel digital beamforming （DBF） algorithm based on the least mean square algorithm （PLMS） is proposed. An appropriate method is found to partition the least mean square （LMS） algorithm into a number of operational modules, which can be easily executed in a distributed-parallel-processing fashion. As a result, the proposed PLMS algorithm provides an effective solution that can alleviate the bottleneck of high-rate data transmission and reduce the computational cost. PLMS requires less computational load than that of the conventional parallel algorithms based on the recursive least square （RLS） algorithm, as well as it is easier to be implemented to do real time adaptive array processing. Moreover, low sidelobe of the beam pattern is obtained by constraining the static steering vector with Tschebyscheff coefficients. Finally, a scheme of the PLMS algorithm using distributed-parallel-processing system is also proposed. The simulation results demonstrate that the PLMS algorithm has the same interference cancellation performance as that of the conventional LMS algorithm. Moreover, the PLMS algorithm can obtain the same good beamforming performance, regardless how the algorithm is partitioned. It is expected that the proposed algorithm will be used in a large-scale adaptive array system to deal with real time adaptive digital beamforming processing.

关键词： adaptive digital beamforming parallel algorithm least mean square generalized sidelobe canceller.

来源：评论

学校读者我要写书评

暂无评论

Surface Reconstruction Technology from Dense Scattered Points Based on Grid

Surface Reconstruction Technology from Dense Scattered Point...

引用

2nd International Conference on High Performance Computing and Applications

作者： Feng, Jianzhou Kong, Lingfu Wang, Xiaohuan Yanshan Univ Sch Informat Sci & Engn Qinhuangdao 066004 Peoples R China

ISBN: (纸本)9783642118418

In order to improve the speed of surface reconstruction from densely scattered points, and reduce the application cost, this paper describes a new and fast surface reconstruction method based on grid computing. The proposed method converts large-scale unorganized 3D scanned datasets into layered datasets firstly. Then based on data parallel mechanism, a loosely coupled parallel reconstruction algorithm is designed;the algorithm has less inter-node communication, so that it is more suitable for grid computing. In order to realize load balance in grid, the priority preemptive scheduling strategy is designed based on two-level scheduling model. Finally, the grid environment is built by Globus Toolkit, and the parallel reconstruction and visualization are achieved based on mpich-G2 and the Visualization Toolkit (VTK), this experiment shows that the reconstruction time is reduced significantly.

关键词： surface reconstruction grid computing parallel algorithm job Scheduling Visualization

来源：评论

学校读者我要写书评

暂无评论

A Distributed Maximum Likelihood algorithm for Multi-Robot Mapping

A Distributed Maximum Likelihood Algorithm for Multi-Robot M...

引用

IEEE/RSJ International Conference on Intelligent Robots and Systems

作者： Rizzini, Dario Lodi Caselli, Stefano Univ Parma Dipartimento Ingn Informaz RIMLab Robot & Intelligent Machines Lab I-43100 Parma Italy

ISBN: (纸本)9781424466757

In the last decade, several algorithms, usually based on information filtering techniques, have been proposed to address multi-robot mapping problem. Less interest has been devoted to investigate a parallel or distributed organization of such algorithms in the perspective of multi-robot exploration. In this paper, we propose a distributed algorithm for map estimation based on Gauss-Seidel relaxation. The complete map is shared among independent tasks running on each robot, which integrate the independent robot measurements in local submaps, and a server, which stores contour nodes separating the submaps. Each task updates its local submap and periodically checks for inter-robot data associations. Gauss-Seidel relaxation is performed independently on each robot and afterwards on the contour nodes set on the server. Results illustrate the potential and flexibility of the new approach.

关键词： Gauss-Seidel relaxation SLAM (robots) cartography distributed algorithm distributed maximum likelihood algorithm information filtering information filtering map estimation maximum likelihood estimation mobile robots multi-robot systems multirobot mapping parallel algorithm parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

On parallelizing Universal Kriging Interpolation based on OpenMP

On Parallelizing Universal Kriging Interpolation based on Op...

引用

9th International Symposium on Distributed Computing and Applications to Business, Engineering and Science (DCABES 2010)

作者： Cheng, Tangpei Li, Dandan Wang, Qun China Univ Geosci Beijing Sch Informat Engn Beijing Peoples R China

ISBN: (纸本)9780769541105

Kriging is one of the important interpolation methods in geostatistics, which has been widely applied in engineering project. In this paper, we present an efficient method for the parallelization of universal Kriging interpolation on shared memory multiprocessors. By using OpenMP directives, we implement a portable parallel algorithm, which enables an incremental approach to add parallelism, without modifying the rest part of sequential code. To achieve optimal performance, the parallel grain size has been considered and analyzed. Numerical experiments have been carried out on two multicore windows workstations, the results of which demonstrate this method could enhance the overall performance significantly.

关键词： Kriging spatial interpolation parallel algorithm OpenMP

来源：评论

学校读者我要写书评

暂无评论

Novel algorithm for simulation of 3D quantum reactive atom-diatom scattering

Novel algorithm for simulation of 3<i>D</i> quantum reactive...

引用

International Conference on Computational Science (ICCS)

作者： Gevorkyan, A. S. Balint-Kurti, G. G. Nyman, G. Inst Informat & Automat Problems NAS Armenia 1 P Sevak Str Yerevan 0014 Armenia Joint Inst Nucl Res Dubna 141980 Russia Univ Bristol Sch Chem Ctr Computat Chem Bristol BS8 1TS Avon England Univ Gothenburg Dept Chem SE-41296 Gothenburg Sweden

A new approach is described to the evaluation of the quantum scattering S-matrix in 3D atom-diatom reactive collision. The theory is developed in terms of natural collision coordinates where the coordinate reaction fulfills the same role as a time in a time-dependent scattering formulation. Having written the full wavefunction of the particles system in the coupled-channel representation we have proved that the 3D multi-channel scattering problem can be reduced to the inelastic single-arrangement problem which is described by system of ordinary differential equations (ODE) of second order. The system of coupled-channel second order ODEs exactly is reduced to the system of integro-differential equations (IDE) of first order which is solved with the initial conditions. The problem of Koshi for the system of IDEs is proposed to be solved by the method of Runge-Kutta of fourth order. The detailed algorithm for parallel simulation of initial 3D scattering problem is proposed. In result of simulation of IDEs the full wavefunction and all S-matrix elements of reactive transitions and state-to-state cross section are obtained simultaneously without other extra calculations. (C) 2010 Published by Elsevier Ltd.

关键词： 3D quantum reactive scattering S-matrix elements coupled-channel differential equation Integro-differential equations Runge-Kutta method parallel algorithm

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：