Metaheuristic algorithms for building Covering Arrays: A review

Covering Arrays (CA) are mathematical objects used in the functional testing of software components. They enable the testing of all interactions of a given size of input parameters in a procedure, function, or logical unit in general, using the minimum number of test cases. Building CA is a complex...

Descripción completa

Detalles Bibliográficos
Autores: Jimena Adriana Timaná-Peña, Carlos Alberto Cobos-Lozada, Jose Torres-Jimenez
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2016
País:México
Institución:Centro de Investigación y de Estudios Avanzados del IPN
Repositorio:Redalyc-CINVESTAV
OAI Identifier:oai:redalyc.org:413948045003
Acceso en línea:https://www.redalyc.org/articulo.oa?id=413948045003
https://www.redalyc.org/journal/4139/413948045003/
https://www.redalyc.org/journal/4139/413948045003/html/
https://www.redalyc.org/journal/4139/413948045003/413948045003.epub
https://www.redalyc.org/journal/4139/413948045003/movil
Access Level:acceso abierto
Palabra clave:Ingeniería
tabu search
metaheuristics
Covering Array
simulated annealing
genetic algorithms
Descripción
Sumario:Covering Arrays (CA) are mathematical objects used in the functional testing of software components. They enable the testing of all interactions of a given size of input parameters in a procedure, function, or logical unit in general, using the minimum number of test cases. Building CA is a complex task that involves lengthy execution times and high computational loads. The most effective methods for building CAs are algebraic, Greedy, and metaheuristic-based. The latter have reported the best results to date. This paper presents a description of the major contributions made by a selection of different metaheuristics, including simulated annealing, tabu search, genetic algorithms, ant colony algorithms, particle swarm algorithms, and harmony search algorithms. It is worth noting that simulated annealing-based algorithms have evolved as the most competitive, and currently form the state of the art.