Heuristics for the robust coloring problem

Abstract

Let G and Ḡ be complementary graphs. Given a penalty function defined on the edges of Ḡ, we will say that the rigidity of a k-coloring of G is the sum of the penalties of the edges of Ḡ joining vertices of the same color. Based on the previous definition, the Robust Coloring Problem (RCP) is stated as the search of the minimum rigidity k-coloring. In this work a comparison of heuristics based on simulated annealing, GRASP and scatter search is presented. These are the best results for the RCP that have been obtained.