A discrete cuckoo optimization algorithm for consolidation in cloud computing. (January 2018)

• Record Type: Journal Article
• Authors: Tavana, Madjid; Shahdi-Pashaki, Saleh; Teymourian, Ehsan; Santos-Arteaga, Francisco J.; Komaki, Mohammad
• Is Part Of: Computers & industrial engineering. Volume 115(2018)
• Page Start: 495
• Page End: 511
• Publication Date: 2018-01
• Subjects: Cloud computing -- Group technology -- Mathematical model -- Virtual machine -- Cuckoo optimization algorithm; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2017.12.001
• Languages: English
• ISSNs: 0360-8352

Abstract

Graphical abstract: Highlights: Group technology (GT) is used to manage resource allocation problems by reducing manufacturing costs. We propose a discrete cuckoo optimization algorithm (DCOA) based on GT for consolidation in cloud computing. We consider grouping strategy using the Jaccard similarity coefficient, egg laying, and immigration of the cuckoos. A numerical example demonstrate the applicability of the proposed model and exhibit the efficacy of the DCOA. The results indicate the superiority of DCOA over the first fit (FF) and round robin (RR) algorithms. Abstract: Consolidation problems in cloud computing (CC) encompass server consolidation, virtual machine (VM) consolidation, and task consolidation. These problems have become increasingly challenging for resource allocation in distributed systems. Group technology (GT) has been effectively used to manage resource allocation problems by reducing manufacturing costs and increasing system productivity. We propose a discrete cuckoo optimization algorithm (DCOA) based on GT for consolidation in CC. The proposed model is designed to control manufacturing costs (i.e., energy, penalty, VM creating, and task migration). The DCOA developed in this study contains several new adjustments that allow it to solve large-sized discrete problems, including a grouping strategy based on the Jaccard similarity coefficient, as well as modified egg laying and immigration processes. A numerical example is used to demonstrate the applicability of the proposed model and exhibit the efficacy of the DCOA. The results illustrate the quality superiority of the DCOA over the first fit (FF) and round robin (RR) algorithms, and the efficiency and effectiveness superiority of the DCOA over the genetic algorithm (GA).