A genetic algorithm for multi-objective optimisation in workflow scheduling with hard constraints. (1st January 2013)

• Record Type: Journal Article
• Authors: Yassa, Sonia; Sublime, Jérémie; Chelouah, Rachid; Kadima, Hubert; Jo, Geun-Sik; Granado, Bertrand
• Is Part Of: International journal of metaheuristics. Volume 2:Number 4(2013)
• Page Start: 415
• Page End: 433
• Publication Date: 2013-01-01
• Subjects: genetic algorithm -- cloud computing -- workflow scheduling -- service level agreement -- quality of service -- QoS -- hard constraints -- metaheuristics; Heuristic algorithms -- Periodicals; 006.3105
• Languages: English
• ISSNs: 1755-2176

Abstract

Cloud computing is a fast growing technology allowing companies to use on-demand computation, and data services for their everyday needs. The main contribution of this work is to propose a new model of genetic algorithm for the workflow scheduling problem. The algorithm must be capable of: 1) dealing with the multi-objective problem of optimising several quality of service (QoS) variables, namely: computation time, cost, reliability or security; 2) handling a large number of workflow scheduling aspects such as adding constraints on QoS variables (deadlines and budgets); 3) handling hard constraints such as restrictions on task scheduling that the previous algorithms have not addressed. Using data from Amazon elastic compute cloud (EC2) and workflows from the London e-Science Centre; we have compared our algorithm with other scheduling algorithms. Simulation results indicate the efficiency of the proposed metaheuristic both in terms of solution quality and computational time.