Application of Parallel and Hybrid Metaheuristics for Graph Partitioning Problem

In this paper parallel and hybrid metaheuristics for graph partitioning are compared taking into account their efficiency in terms of a cost function and computation time. Seventeen methods developed on the basis of evolutionary algorithm, simulated annealing and tabu search are implemented and test...

Full description

Saved in:
Bibliographic Details
Published inNumerical Methods and Applications Vol. 11189; pp. 125 - 132
Main Authors Kokosiński, Zbigniew, Pijanowski, Marcin
Format Book Chapter
LanguageEnglish
Published Switzerland Springer International Publishing AG 2019
Springer International Publishing
SeriesLecture Notes in Computer Science
Subjects
Approximate algorithms
Circuit partitioning
DIMACS graphs
Graph partitioning
Hybrid metaheuristics
Parallel metaheuristics
R–MAT graphs
Online AccessGet full text
ISBN9783030106911
3030106918
ISSN0302-9743
1611-3349
DOI10.1007/978-3-030-10692-8_14

Cover

More Information
Summary:In this paper parallel and hybrid metaheuristics for graph partitioning are compared taking into account their efficiency in terms of a cost function and computation time. Seventeen methods developed on the basis of evolutionary algorithm, simulated annealing and tabu search are implemented and tested against graph instances computed on the basis of queen graphs from DIMACS repository and a class of random R–MAT graphs. These graphs are supposed to model a class of digital circuits being subject of decomposition into a given number of modules. In partitioning process several additional constraints have to be satisfied in order to enable composition of original circuits from subcircuits by means of VLSI/FPGA modules.
ISBN:9783030106911
3030106918
ISSN:0302-9743
1611-3349
DOI:10.1007/978-3-030-10692-8_14