Title: The improved ColourAnt algorithm: a hybrid algorithm for solving the graph colouring problem

Authors: Anderson Faustino Da Silva; Luis Gustavo Araujo Rodriguez; João Fabrício Filho

Addresses: Department of Informatics, State University of Maringá, Maringá, PR, Brazil ' Graduate School in Computer Science, University of São Paulo, São Paulo, SP, Brazil ' Federal University of Technology, Campo Mourão, PR, Brazil

Abstract: The graph colouring problem is interesting because of its application areas, ranging from register allocation, frequency association in telecommunications, timetabling and scheduling, and others. This problem is NP-complete and thus, several metaheuristic algorithms have been proposed in order to provide a good solution in an acceptable time-frame. Among several metaheuristics, this paper focuses on ant colony optimisation. In this context, a hybrid algorithm was developed called iColourAnt, which uses ant colony optimisation and an efficient local search strategy, consequently providing good solutions to the graph colouring problem. The experiment results indicate that iColourAnt outperforms its predecessor, ColourAnt.

Keywords: graph colouring; metaheuristic; ant colony optimisation; ACO.

DOI: 10.1504/IJBIC.2020.109000

International Journal of Bio-Inspired Computation, 2020 Vol.16 No.1, pp.1 - 12

Accepted: 29 Jan 2020
Published online: 14 Aug 2020 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article