Metaheuristics for scheduling of heterogeneous tasks in cloud computing environments: Analysis, performance evaluation, and future directions

Abstract In cloud computing environments, when a client wants to access any resources, hardware components, or application services, he needs to get a subscription for the same from service providers. The usages of each client are monitored over a network by service providers and later on user will be charged for the services used. Cloud service provider is responsible for providing Quality of Service to clients. As the number of client request increases in cloud environment, cloud service providers face various issues such as scheduling and allocation of resources, security, privacy and virtual machine migration. Swarm intelligence, biological systems, physical and chemical systems based metaheuristic algorithms have proved to be efficient and used to solve real world scheduling optimization problems. This review focused on the insight view of various nature-inspired metaheuristic algorithms and their comparisons on the basis of certain parameters that affects the efficiency and effectiveness of their applicability in order to schedule different tasks in cloud environment. This work facilitates comparative analysis of six metaheuristic techniques quantitatively based on scheduling parameters like makespan and resource utilization cost. The objective of this systematic review is to find the most optimal scheduling technique for solving multi criteria scheduling problem. After evaluating and comparing Ant Colony Optimization, Particle Swarm Optimization, Genetic Algorithm, Artificial Bee Colony algorithm, Crow Search Algorithm and Penguin Swarm Optimization Algorithm, it has been identified that Crow Search algorithm is the most optimal technique in terms of makespan and resource utilization cost parameters with significant improvement over others. Finally, the promising research directions has been identified.