A reactive search optimization algorithm for scientific workflow scheduling using clustering techniques

摘要

Cloud computing is the style that can give plenty of shared pool resources such as hardware or software to clients based on requests from the internet. These resources are then scaled up automatically based on the specifications of the clients. Workflow scheduling optimization is an area of research activities in infrastructure as a service (IaaS) of the cloud. This problem is NP-complete. Thus, building a workflow scheduler that is optimum, having a reasonable level of performance and speed of computation, can be quite challenging in a distributed cloud environment. Metaheuristic algorithms may be improved in terms of their solution and its quality and speed of convergence utilizing combining it with other metaheuristic algorithms or any other algorithms that are metaheuristic based on local search. Shuffled frog leaping algorithm (SFLA) was acknowledged a metaheuristic performing heuristic search with a heuristic function (mathematical function) seeking solutions to combinatorial optimization problems. An optimization ratio on makespan %, resource utilization and computational cost performs better for SFLA-RSO with clustering when the number of tasks are increased.