|
1. |
SYMMETRIC DOMAIN DECOMPOSITION METHOD FOR LARGE SCALE AND NONLINEAR EVOLUTION SYSTEMS* |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 161-175
QIMING HE,
DAVIDJ. EVANS,
Preview
|
PDF (205KB)
|
|
摘要:
The domain decomposition method is one of the most important and effective methods to obtain the numerical solutions for partial differential equations at present. And the symmetric domain decomposition method, although a special domain decomposition method just developed recently, has very extensive application prospects shown by some present work. The purpose of this article is to generalize this method and apply it to large scale and nonlinear evolution systems. It is demonstrated by theoretical analyses that such method is stable and convergent. And an example is employed to show the applicability and efficiency of this method.
ISSN:1063-7192
DOI:10.1080/10637199508915529
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
2. |
MODULAR FIXED-SIZE VLSI ARCHITECTURES FOR GENERAL MULTISPLITTING ITERATION |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 177-191
ELENA PAPADOPOULOU,
YIANNIS SARIDAKIS,
Preview
|
PDF (226KB)
|
|
摘要:
Motivated by the inherent parallelicity of the Multisplitting Iterative Methods, we consider their application for the solution of Large Linear Systems. The realistic parallel implementation of this problem led us to the employment of fixed-size VLSI architectures. Considering the case of General Splitting matrices we combine known as well as we design new VLSI BLAS modules to form fixed-size architectures capable of efficiently carrying out the computations involved in an oversized multisplitting iteration step. For the factorization of the splitting matrices we employ the LU decomposition method while for the organization of the data streams we use space-time partitioning techniques.
ISSN:1063-7192
DOI:10.1080/10637199508915530
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
3. |
THE PARALLEL ALGORITHMS FOR DETERMINING EDGE-PACKING AND EFFICIENT EDGE DOMINATING SETS IN INTERVAL GRAPHS |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 193-207
MADHUMANGAL PAL,
G. P. BHATTACHARJEE,
Preview
|
PDF (275KB)
|
|
摘要:
Recently, it has been shown that resource allocation problems in parallel processing systems can be viewed as edge domination problems in graphs. Other applications of edge domination include encoding theory and network routing problems. In a graph G = (V,E) and edge (u,v) ∈ E is said to dominate itself and any edge (u,x) or (v,x) where x ∈ V. An Edge-Packing (EP) in a graph G is a set of edges (B), B CE such that no edge in E is dominated by more than one edge of B. A subset of edges E’ C E is called an Efficient Edge Domination (EED) set for the graph G if all edges in E are dominated by exactly one edge of E’. The EED problem for general graph is NP-complete. For the series parallel graph a linear time sequential algorithm is available. In this paper, a linear time sequential algorithm is presented to find EP for a weighted interval graph. Parallel algorithms are also presented to find EP for weighted and unweighted interval graphs. For the weighted case, the proposed parallel algorithm takes O(log2n) time and O(n3/logn) processors and for the unweighted case, the parallel algorithm lakes O(logn) time and O(n + m) processors on an EREW PRAM, where m,n represent number of edges and number of vertices of the graph. If EED set exists for the given unweighted interval graph then it can be computed using the same resource bound for finding EP of an unweighted interval graph.
ISSN:1063-7192
DOI:10.1080/10637199508915531
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
4. |
PARSING UNCERTAIN CONTEXT-FREE LANGUAGES BY GENETIC ALGORITHMS |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 209-228
L. BARSAN,
C. BARSAN,
D. J. EVANS,
Preview
|
PDF (289KB)
|
|
摘要:
The parsing of context free languages can be done by efficient algorithms inO(n3) for general context-free case and inO(n) for certain classes as LL and LR. All these algorithms are very strict, they fail to recover on any insignificant input error. In this paper we propose a robust parsing technique for uncertain inputs based on genetic algorithms. As the genetic techniques are computational expensive the genetic parser is not intended to substitute classical parser, but to assist it. During the parsing process, erroneous input words could be passed to the genetic parser in order to find the best approximation of the input word.
ISSN:1063-7192
DOI:10.1080/10637199508915532
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
5. |
THE CONVERGENCE OF ASYNCHRONOUS ITERATIONS FOR THE FIXED POINT OF A SPLITTING OPERATOR |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 229-235
LEI LI,
TADAO NAKAMURA,
Preview
|
PDF (91KB)
|
|
摘要:
In this paper, we present the convergence of asynchronous iterations for the fixed point of the splitting operator using Baudet' asynchronous iteration method on multiprocessors.
ISSN:1063-7192
DOI:10.1080/10637199508915533
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
6. |
PARALLEL UPDATING TECHNIQUES IN IMAGE RESTORATION |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 237-253
P. BLAZNIK,
D. J. EVANS,
J. TASIČ,
Preview
|
PDF (203KB)
|
|
摘要:
In this paper, the linear algebra techniques used in image restoration are outlined, and updating algorithms on a systolic array model are presented for solving the resulting linear systems of equations.
ISSN:1063-7192
DOI:10.1080/10637199508915534
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
7. |
A CROSS-PLATFORM PARALLEL GENETIC ALGORITHMS PROGRAMMING ENVIRONMENT |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 255-266
K. TOUT,
J. RIBEIRO-FILHO,
B. MIGNOT,
Preview
|
PDF (219KB)
|
|
摘要:
Genetic algorithms (GAs) have been widely and successfully used for solving complex optimisation problems. This paper presents the research being conducted under the ESPRIT III PAPAGENA project which aims to establish an European standard for the design of adaptive genetic systems. It describes the architecture of the Genetic Algorithms Manipulation Environment (GAME). GAME is, a parallel, general-purpose object-oriented programming environment that offers extensive tools for the design, configuration and monitoring of Parallel Genetic Algorithm-based (PGA) applications in a variety of domains. To exploit the intrinsic parallelism of GAs, two parallelisation options are available within GAME: implicit, where the parallelism is low-level and transparent to the application developer; andexplicitin which the parallelism is managed by the application developer, and could be introduced at different levels and in different ways. The parallel layer of GAME is hardware platform-independent. This property means that it is portable on a set of currently available parallel platforms. A special consideration has been given for the implementation on parallel architectures based on T800 and T9000 transputers.
ISSN:1063-7192
DOI:10.1080/10637199508915535
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
8. |
PERFORMANCE EVALUATION OF BACKPROPAGATION NEURAL NETWORK IMPLEMENTATION ON SINGLE PROGRAM MULTIPLE DATA (SPMD) ARCHITECTURE |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 267-281
AKHIL SAHAI,
L. M. PATNAIK,
Preview
|
PDF (230KB)
|
|
摘要:
The Single Program Multiple Data (SPMD) approach has been developed to reduce the system overheads, to entail minimal computation time and to reduce the programming complexity of parallel programs. The SPMD computational model is based on the premise that all the processes execute the same program; however, at any instant of time, these processes may be executing different instructions and may be operating on different data. This work uses the SPMD computational model to implement a backpropagation neural network and the implementation is done on a multiprocessor. We present the distributed backpropagation forward and backward execution algorithms in SPMD mode. The performance of the SPMD-implemented ANN is demonstrated by the Automatic Target Recognition (ATR) problem, which involves identification of a class of targets depending on a set of input data values. The performance of the SPMD-implemented ANN is studied by observing the speedups achieved for training times as the number of processors is increased. The saving in training times is found to be quite impressive. The ATR problem up to 3 classes of targets has been tried and the ANN is found to perform the recognition easily.
ISSN:1063-7192
DOI:10.1080/10637199508915536
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
9. |
A COMPARISON STUDY OF HEURISTICS FOR MAPPING PARALLEL ALGORITHMS TO MESSAGE-PASSING MULTIPROCESSORS |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 283-302
MARIOSD. DIKAIAKOS,
ANNE ROGERS,
KEN STEIGLITZ,
Preview
|
PDF (376KB)
|
|
摘要:
This paper presents a comparison study of popular clustering and mapping heuristics which are used to map task-flow graphs to message-passing multiprocessors. Tb this end, we use task-graphs which arc representative of important scientific algorithms running on data-sets of practical interest. The annotation which assigns weights to nodes and edges of the task-graphs is realistic It reflects current trends in processor, communication channel, and message-passing interface technology and takes into consideration hardware characteristics of state-of-the-art multiprocessors. Our experiments show that applying realistic models for task-graph annotation affects the effectiveness and functionality of clustering and mapping techniques. Therefore, new heuristics are necessary that will take into account more practical models of communication costs. We present modifications to existing clustering and mapping algorithms which improve their efficiency and running-time for the practical models adopted.
ISSN:1063-7192
DOI:10.1080/10637199508915537
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
10. |
A UNIVERSAL MAPPING FOR NESTED LOOPS IN PARALLEL PROGRAMMING |
|
Parallel Algorithms and Applications,
Volume 7,
Issue 3-4,
1995,
Page 303-311
D. J. EVANS,
N. M. BAHOSHY,
Preview
|
PDF (120KB)
|
|
摘要:
In this paper a general method for partitioning arbitrary nested loops in parallel programming is described and implemented.
ISSN:1063-7192
DOI:10.1080/10637199508915538
出版商:Taylor & Francis Group
年代:1995
数据来源: Taylor
|
|