Volume 18, No. 6, 2021
Optimization Of Tasks Scheduling In Computational Grids Using Hybrid Swarm Intelligence Algorithm
Kishor Shamrao Sakure , Rajesh Kumar Boghey
Abstract
The scalability and reliability of computational grids are challenging tasks in a next-generation computational framework. The growing demands of distributed resources face a problem of allocation and decline the service quality of computational grids. The optimization of resources is a way to handle the problem of computational grids. This paper proposed a hybrid swarm intelligence-based task scheduling algorithm for allocating resources in computational grids. The proposed algorithm encapsulates two algorithms ant colony optimization and particle swarm optimization. The ant colony optimization algorithm handles the global request of resources, and particle swarm optimization handles the resource's local searching. The hybrid swarm intelligence algorithm scales the resource allocation process and reduces the job failure rate instead of other swarm intelligence algorithms. The proposed algorithm was simulated in MATLAB tools and tested three sizes of grid matrices. The size of the grid matrix varies from lower to higher and increase the load in respect of resource. The proposed hybrid swarm intelligence algorithm compares with the existing resource optimization algorithm. The analysis of results suggests that the proposed algorithm increase 2-3 % of job completion instead of the existing algorithm.
Pages: 3711-3725
Keywords: Grid, High-Performance Computing, Swarm Intelligence, Optimization, Scheduling, PSO, ACO