Fast parallel simulated annealing for traveling salesman problem

Chang-Sung Jeong, M. H. Kim

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

9 Citations (Scopus)

Abstract

The authors present a fast parallel simulated annealing (SA) algorithm for solving the traveling salesman problem (TSP) on single instruction stream, multiple data stream (SIMD) machines with linear interconnections among processing elements. The whole simulated annealing algorithm has been sped up by implementing the move operation in parallel. A generation scheme which can be efficiently adapted to parallel implementation and achieve the balanced permutation has been designed. It is shown that, with the proper data distribution and movement schemes, the generation of a new configuration and the calculation of energy difference can be done in constant time, and the whole time complexity of the move operation is proportional to the time taken to broadcast 1 b of information on acceptance decision to all the other PEs (processing elements) on a linear class computer with the same number of PEs as that of cities. Therefore, if the control unit has the capability of broadcasting 1 b in one PE to all the other PEs, as is often the case with the SIMD machine, the time complexity of the whole SA algorithm can be reduced to the order of the number of the moves.

Original languageEnglish
Title of host publicationIJCNN. International Joint Conference on Neural Networks
Place of PublicationPiscataway, NJ, United States
PublisherPubl by IEEE
Pages947-953
Number of pages7
Publication statusPublished - 1990 Dec 1
Externally publishedYes
Event1990 International Joint Conference on Neural Networks - IJCNN 90 Part 3 (of 3) - San Diego, CA, USA
Duration: 1990 Jun 171990 Jun 21

Other

Other1990 International Joint Conference on Neural Networks - IJCNN 90 Part 3 (of 3)
CitySan Diego, CA, USA
Period90/6/1790/6/21

Fingerprint

Traveling salesman problem
Simulated annealing
Processing
Broadcasting

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Jeong, C-S., & Kim, M. H. (1990). Fast parallel simulated annealing for traveling salesman problem. In IJCNN. International Joint Conference on Neural Networks (pp. 947-953). Piscataway, NJ, United States: Publ by IEEE.

Fast parallel simulated annealing for traveling salesman problem. / Jeong, Chang-Sung; Kim, M. H.

IJCNN. International Joint Conference on Neural Networks. Piscataway, NJ, United States : Publ by IEEE, 1990. p. 947-953.

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

Jeong, C-S & Kim, MH 1990, Fast parallel simulated annealing for traveling salesman problem. in IJCNN. International Joint Conference on Neural Networks. Publ by IEEE, Piscataway, NJ, United States, pp. 947-953, 1990 International Joint Conference on Neural Networks - IJCNN 90 Part 3 (of 3), San Diego, CA, USA, 90/6/17.
Jeong C-S, Kim MH. Fast parallel simulated annealing for traveling salesman problem. In IJCNN. International Joint Conference on Neural Networks. Piscataway, NJ, United States: Publ by IEEE. 1990. p. 947-953
Jeong, Chang-Sung ; Kim, M. H. / Fast parallel simulated annealing for traveling salesman problem. IJCNN. International Joint Conference on Neural Networks. Piscataway, NJ, United States : Publ by IEEE, 1990. pp. 947-953
@inproceedings{52359528d80e42e1bf62770aa007b36a,
title = "Fast parallel simulated annealing for traveling salesman problem",
abstract = "The authors present a fast parallel simulated annealing (SA) algorithm for solving the traveling salesman problem (TSP) on single instruction stream, multiple data stream (SIMD) machines with linear interconnections among processing elements. The whole simulated annealing algorithm has been sped up by implementing the move operation in parallel. A generation scheme which can be efficiently adapted to parallel implementation and achieve the balanced permutation has been designed. It is shown that, with the proper data distribution and movement schemes, the generation of a new configuration and the calculation of energy difference can be done in constant time, and the whole time complexity of the move operation is proportional to the time taken to broadcast 1 b of information on acceptance decision to all the other PEs (processing elements) on a linear class computer with the same number of PEs as that of cities. Therefore, if the control unit has the capability of broadcasting 1 b in one PE to all the other PEs, as is often the case with the SIMD machine, the time complexity of the whole SA algorithm can be reduced to the order of the number of the moves.",
author = "Chang-Sung Jeong and Kim, {M. H.}",
year = "1990",
month = "12",
day = "1",
language = "English",
pages = "947--953",
booktitle = "IJCNN. International Joint Conference on Neural Networks",
publisher = "Publ by IEEE",

}

TY - GEN

T1 - Fast parallel simulated annealing for traveling salesman problem

AU - Jeong, Chang-Sung

AU - Kim, M. H.

PY - 1990/12/1

Y1 - 1990/12/1

N2 - The authors present a fast parallel simulated annealing (SA) algorithm for solving the traveling salesman problem (TSP) on single instruction stream, multiple data stream (SIMD) machines with linear interconnections among processing elements. The whole simulated annealing algorithm has been sped up by implementing the move operation in parallel. A generation scheme which can be efficiently adapted to parallel implementation and achieve the balanced permutation has been designed. It is shown that, with the proper data distribution and movement schemes, the generation of a new configuration and the calculation of energy difference can be done in constant time, and the whole time complexity of the move operation is proportional to the time taken to broadcast 1 b of information on acceptance decision to all the other PEs (processing elements) on a linear class computer with the same number of PEs as that of cities. Therefore, if the control unit has the capability of broadcasting 1 b in one PE to all the other PEs, as is often the case with the SIMD machine, the time complexity of the whole SA algorithm can be reduced to the order of the number of the moves.

AB - The authors present a fast parallel simulated annealing (SA) algorithm for solving the traveling salesman problem (TSP) on single instruction stream, multiple data stream (SIMD) machines with linear interconnections among processing elements. The whole simulated annealing algorithm has been sped up by implementing the move operation in parallel. A generation scheme which can be efficiently adapted to parallel implementation and achieve the balanced permutation has been designed. It is shown that, with the proper data distribution and movement schemes, the generation of a new configuration and the calculation of energy difference can be done in constant time, and the whole time complexity of the move operation is proportional to the time taken to broadcast 1 b of information on acceptance decision to all the other PEs (processing elements) on a linear class computer with the same number of PEs as that of cities. Therefore, if the control unit has the capability of broadcasting 1 b in one PE to all the other PEs, as is often the case with the SIMD machine, the time complexity of the whole SA algorithm can be reduced to the order of the number of the moves.

UR - http://www.scopus.com/inward/record.url?scp=0025532284&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0025532284&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0025532284

SP - 947

EP - 953

BT - IJCNN. International Joint Conference on Neural Networks

PB - Publ by IEEE

CY - Piscataway, NJ, United States

ER -