A discrete heuristic algorithm with swarm and evolutionary features for data replication problem in distributed systems

Basit Öğe Kaydı
dc.authoridArasteh, Bahman/0000-0001-5202-6315
dc.authoridAllahviranloo, Tofigh/0000-0002-6673-3560
dc.authoridCatak, Muammer/0000-0003-0752-2525
dc.authorwosidArasteh, Bahman/AAN-9555-2021
dc.authorwosidAllahviranloo, Tofigh/V-4843-2019
dc.authorwosidKhari, Manju/B-6040-2017
dc.authorwosidTorkamanian-Afshar, Mahsa/AAD-9989-2022
dc.contributor.authorArasteh, Bahman
dc.contributor.authorAllahviranloo, Tofigh
dc.contributor.authorFunes, Peri
dc.contributor.authorTorkamanian-Afshar, Mahsa
dc.contributor.authorKhari, Manju
dc.contributor.authorCatak, Muammer
dc.date.accessioned2024-05-19T14:40:55Z
dc.date.available2024-05-19T14:40:55Z
dc.date.issued2023
dc.departmentİstinye Üniversitesien_US
dc.description.abstractAvailability and accessibility of data objects in a reasonable time is a main issue in distributed systems like cloud computing services. As a result, the reduction of data-related operation times in distributed systems such as data read/write has become a major challenge in the development of these systems. In this regard, replicating the data objects on different servers is one commonly used technique. In general, replica placement plays an essential role in the efficiency of distributed systems and can be implemented statically or dynamically. Estimation of the minimum number of data replicas and the optimal placement of the replicas is an NP-complete optimization problem. Hence, different heuristic algorithms have been proposed for optimal replica placement in distributed systems. Reducing data processing costs as well as the number of replicas, and increasing the reliability of the replica placement algorithms are the main goals of this research. This paper presents a discrete and swarm-evolutionary method using a combination of shuffle-frog leaping and genetic algorithms to data-replica placement problems in distributed systems. The experiments on the standard dataset show that the proposed method reduces data access time by up to 30% with about 14 replicas; whereas the generated replicas by the GA and ACO are, respectively, 24 and 30. The average reduction in data access time by GA and ACO 21% and 18% which shows less efficiency than the SFLA-GA algorithm. Regarding the results, the SFLA-GA converges on the optimal solution before the 10th iteration, which shows the higher performance of the proposed method. Furthermore, the standard deviation among the results obtained by the proposed method on several runs is about 0.029, which is lower than other algorithms. Additionally, the proposed method has a higher success rate than other algorithms in the replica placement problem.en_US
dc.identifier.doi10.1007/s00521-023-08853-x
dc.identifier.endpage23197en_US
dc.identifier.issn0941-0643
dc.identifier.issn1433-3058
dc.identifier.issue31en_US
dc.identifier.scopus2-s2.0-85169163792en_US
dc.identifier.scopusqualityQ1en_US
dc.identifier.startpage23177en_US
dc.identifier.urihttps://doi.org10.1007/s00521-023-08853-x
dc.identifier.urihttps://hdl.handle.net/20.500.12713/5036
dc.identifier.volume35en_US
dc.identifier.wosWOS:001062036000002en_US
dc.identifier.wosqualityN/Aen_US
dc.indekslendigikaynakWeb of Scienceen_US
dc.indekslendigikaynakScopusen_US
dc.language.isoenen_US
dc.publisherSpringer London Ltden_US
dc.relation.ispartofNeural Computing & Applicationsen_US
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanıen_US
dc.rightsinfo:eu-repo/semantics/closedAccessen_US
dc.snmz20240519_kaen_US
dc.subjectDistributed Systemsen_US
dc.subjectData Access Timeen_US
dc.subjectReplica Placementen_US
dc.subjectSfla-Ga Optimizationen_US
dc.subjectConvergence Speeden_US
dc.subjectNumber Of Replicasen_US
dc.titleA discrete heuristic algorithm with swarm and evolutionary features for data replication problem in distributed systemsen_US
dc.typeArticleen_US

Dosyalar

Koleksiyon

WoS İndeksli Yayınlar Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu