Open AccessANT COLONY OPTIMIZATION BASED WORKFLOW SCHEDULING IN CLOUD COMPUTING
Author(s) -
Deepika Saxena,
R.K. Chauhan
Publication year - 2017
Publication title -
zenodo (cern european organization for nuclear research)
Language(s) - English
DOI - 10.5281/zenodo.229903
Subject(s) - ant colony optimization algorithms , workflow , computer science , cloud computing , scheduling (production processes) , distributed computing , workflow management system , operating system , database , artificial intelligence , mathematical optimization , mathematics
In the present scenario of Information and Technology, Cloud Computing has become buzzword. Here, dynamically scalable services and distributed virtualized resources are provided over the internet on pay-as-per use basis. Instantaneously, there are huge numbers of users accessing services of cloud and various tasks need to be handled in the cloud computing environment, the high effective task scheduling algorithm is one of the crucial problems that the cloud computing is required to solve. Cloud task scheduling is an NP-hard optimization problem and many different meta-heuristic algorithms have been proposed to solve it. A good task scheduler should adapt its scheduling strategy dynamically according to the changing environment and the types of tasks. Aiming to the model structure of cloud computing, in this article we have introduced modified Ant Colony Optimization algorithm (ACO) to combine with optimized task scheduling algorithm which is dynamic and adapt according to the availability of resources. This paper relates advanced heuristic and combinatorial optimization problem solving technique i.e. Ant Colony Optimization (ACO) which outperforms over other evolutionary algorithm and optimization technique. In proposed algorithm, group of tasks are represented as workflow are scheduled by ants based on heuristic function to the virtual machine. This means all the available tasks are efficiently scheduled to the very best of its optimization. We recompile the cloudsim and simulate the proposed algorithm and results of this algorithm are compared with sequential task scheduling. The experimental results indicates that proposed algorithm has high performance in terms of least execution time that considers heterogeneous resources and elasticity of clouds that can be dynamically acquired on pay-per-use basis. This algorithm is not only beneficial to user and service provider, but also provides better efficiency by applying load-balancing feature i.e. benefit at system level