Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ RUC. Repositorio da ...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
Recolector de Ciencia Abierta, RECOLECTA
Bachelor thesis . 2022
License: CC BY NC ND
image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao
Recolector de Ciencia Abierta, RECOLECTA
Bachelor thesis . 2022
License: CC BY NC ND
versions View all 4 versions
addClaim

Estratexias de tolerancia a fallos nun algoritmo paralelo de colonia de formigas

Authors: Blanco Godón, Miguel;

Estratexias de tolerancia a fallos nun algoritmo paralelo de colonia de formigas

Abstract

[Resumo]: A resolución de problemas combinatorios de complexidade NP aínda é un problema aberto debido ao alto tempo de computación requirido para obter solucións de boa calidade. Dentro deste conxunto, o problema do viaxante - Travel Salesman Problem (TSP) - é un dos máis estudados pola comunidade científica debido á grande cantidade de aplicacións no mundo real que podemos atopar. Co obxectivo de obter solucións óptimas de xeito eficiente, unha das estratexias máis prometedoras é o uso de algoritmos de colonia de formigas - Ant Colony Optimization (ACO) - e variacións deste baseados no uso de metaheurísticas. Non tardaron en aparecer solucións paralelas para este algoritmo para intentar acelerar a computación da solución aínda máis. Non obstante, o aumento do número de recursos utilizados nas aplicacións e sistemas paralelos, conleva ao aumento na taxa de fallos aos que están expostos estas aplicacións tan demandantes, polo que utilizar técnicas de tolerancia a fallos vólvese fundamental. O obxectivo deste proxecto é estudar diferentes estratexias de tolerancia a fallos para unha versión paralela asíncrona do ACO, utilizando as extensións de MPI que proporciona o framework User Level Failure Mitigation (ULFM). Esta proposta foi avaliada utilizando dous problemas TSP distintos extraídos da libraría TSPLib, obtendo tolerancia dende 1 a N − 1 fallos (sendo N o número de procesos MPI) cun overhead moi pequeno na maior parte dos casos.

[Abstract]: Solving combinatorial problems of NP complexity is still an open problem due to the high computational time required to obtain good quality solutions. Within this set, Travel Salesman Problem (TSP) is one of the most studied by the scientific community due to the large number of applications in the real world that we can find. With the aim of obtaining optimal solutions efficiently, one of the most promising strategies is the use of ant colony algorithms - Ant Colony Optimization (ACO) - and variations of it based on the use of metaheuristics. Parallel solutions to this algorithm soon appeared in an attempt to speed up the computation of the solution even more. However, the increase in the number of resources used in parallel applications and systems leads to an increase in the rate of failures to which these demanding applications are exposed, so using fault tolerance techniques becomes essential. The objective of this project is to study different fault tolerance strategies for an asynchronous parallel version of the ACO, using the MPI extensions provided by the User Level Failure Mitigation (ULFM) framework. This proposal was evaluated using two TSP problems taken from the TSPLib library, obtaining tolerance from 1 to N − 1 failures (being N the number of MPI processes) with a very small overhead in most cases.

Traballo fin de grao (UDC.FIC). Enxeñaría Informática. Curso 2021/2022

Country
Spain
Related Organizations
Keywords

ACO, Parallel computing, Computación paralela, Fault tolerance, MPI, Tolerancia a fallos, ULFM

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green