Dynamic iterative method for fast network partitioning

Chang-Sung Jeong, Young Min Song, Sung Up Jo

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

In this paper, we address multiway network partitioning problem of dividing the cells of network into multiple blocks so as to minimize the number of nets interconnecting cells in different blocks while balancing the blocks' sizes. The sequential iterative improvement algorithm for the problem consists of several passes each of which is performed by repeatedly iterating the move operation. Therefore, the whole execution time taken by the algorithm is greatly affected by the number of the move operations and the execution time for each move operation. We present a fast parallel algorithm for solving the multiway network partiotioning problem by reducing both of them. We propose a new dynamic iterative method which reduces the number of move operations executed at each pass dynamically, and hence speed up the whole algorithm sharply. Moreover, we reduced the execution time of each move by its parallelization using the proper cell distribution.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages81-88
Number of pages8
Volume1823
ISBN (Print)9783540675532
Publication statusPublished - 2000
Event8th International Conference on High Performance Computing and Networking, HPCNEurope 2000 - Amsterdam, Netherlands
Duration: 2000 May 82000 May 10

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1823
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other8th International Conference on High Performance Computing and Networking, HPCNEurope 2000
CountryNetherlands
CityAmsterdam
Period00/5/800/5/10

Fingerprint

Iterative methods
Partitioning
Execution Time
Iteration
Cell
Parallel algorithms
Parallelization
Balancing
Parallel Algorithms
Fast Algorithm
Speedup
Minimise

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Jeong, C-S., Song, Y. M., & Jo, S. U. (2000). Dynamic iterative method for fast network partitioning. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1823, pp. 81-88). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1823). Springer Verlag.

Dynamic iterative method for fast network partitioning. / Jeong, Chang-Sung; Song, Young Min; Jo, Sung Up.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1823 Springer Verlag, 2000. p. 81-88 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1823).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Jeong, C-S, Song, YM & Jo, SU 2000, Dynamic iterative method for fast network partitioning. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 1823, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1823, Springer Verlag, pp. 81-88, 8th International Conference on High Performance Computing and Networking, HPCNEurope 2000, Amsterdam, Netherlands, 00/5/8.
Jeong C-S, Song YM, Jo SU. Dynamic iterative method for fast network partitioning. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1823. Springer Verlag. 2000. p. 81-88. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Jeong, Chang-Sung ; Song, Young Min ; Jo, Sung Up. / Dynamic iterative method for fast network partitioning. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1823 Springer Verlag, 2000. pp. 81-88 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{0fe41917da364d9fab9d2eb54ca728a3,
title = "Dynamic iterative method for fast network partitioning",
abstract = "In this paper, we address multiway network partitioning problem of dividing the cells of network into multiple blocks so as to minimize the number of nets interconnecting cells in different blocks while balancing the blocks' sizes. The sequential iterative improvement algorithm for the problem consists of several passes each of which is performed by repeatedly iterating the move operation. Therefore, the whole execution time taken by the algorithm is greatly affected by the number of the move operations and the execution time for each move operation. We present a fast parallel algorithm for solving the multiway network partiotioning problem by reducing both of them. We propose a new dynamic iterative method which reduces the number of move operations executed at each pass dynamically, and hence speed up the whole algorithm sharply. Moreover, we reduced the execution time of each move by its parallelization using the proper cell distribution.",
author = "Chang-Sung Jeong and Song, {Young Min} and Jo, {Sung Up}",
year = "2000",
language = "English",
isbn = "9783540675532",
volume = "1823",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "81--88",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Dynamic iterative method for fast network partitioning

AU - Jeong, Chang-Sung

AU - Song, Young Min

AU - Jo, Sung Up

PY - 2000

Y1 - 2000

N2 - In this paper, we address multiway network partitioning problem of dividing the cells of network into multiple blocks so as to minimize the number of nets interconnecting cells in different blocks while balancing the blocks' sizes. The sequential iterative improvement algorithm for the problem consists of several passes each of which is performed by repeatedly iterating the move operation. Therefore, the whole execution time taken by the algorithm is greatly affected by the number of the move operations and the execution time for each move operation. We present a fast parallel algorithm for solving the multiway network partiotioning problem by reducing both of them. We propose a new dynamic iterative method which reduces the number of move operations executed at each pass dynamically, and hence speed up the whole algorithm sharply. Moreover, we reduced the execution time of each move by its parallelization using the proper cell distribution.

AB - In this paper, we address multiway network partitioning problem of dividing the cells of network into multiple blocks so as to minimize the number of nets interconnecting cells in different blocks while balancing the blocks' sizes. The sequential iterative improvement algorithm for the problem consists of several passes each of which is performed by repeatedly iterating the move operation. Therefore, the whole execution time taken by the algorithm is greatly affected by the number of the move operations and the execution time for each move operation. We present a fast parallel algorithm for solving the multiway network partiotioning problem by reducing both of them. We propose a new dynamic iterative method which reduces the number of move operations executed at each pass dynamically, and hence speed up the whole algorithm sharply. Moreover, we reduced the execution time of each move by its parallelization using the proper cell distribution.

UR - http://www.scopus.com/inward/record.url?scp=84944048003&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84944048003&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84944048003

SN - 9783540675532

VL - 1823

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 81

EP - 88

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Verlag

ER -