ICTACT Journal on Soft Computing

P. Mathiyalagan ¹, U. R. Dhepthie ¹, S. N. Sivanandam ²

Affiliations
1 Department of Computer Science and Engineering, PSG College of Technology, Tamil Nadu, IN
2 Akshaya College of Engineering and Technology, Tamil Nadu, IN

Abstract

Grid computing is a high performance computing used to solve larger scale computational demands. Task scheduling is a major issue in grid computing systems. Scheduling of tasks is the NP hard problem. The heuristic approach provides optimal solution for NP hard problems .The ant colony algorithm provides optimal solution. The existing ant colony algorithm takes more time to schedule the tasks. In this paper ant colony algorithm improved by enhancing pheromone updating rule such that it schedules the tasks efficiently and better resource utilization. The simulation results prove that proposed method reduces the execution time of tasks compared to existing ant colony algorithm.

Keywords

Pheromone, Swarm Intelligence, Inertia, Grid Scheduling.

Full Text

     

P. Mathiyalagan ¹, U. R. Dhepthie ¹, S. N. Sivanandam ¹

Affiliations
1 Department of Computer Science and Engineering, PSG College of Technology, Coimbatore, Tamil Nadu, IN

Abstract

Grid computing is a high performance computing environment to solve larger scale computational demands. Grid computing contains resource management, task scheduling, security problems, information management and so on. Task scheduling is a fundamental issue in achieving high performance in grid computing systems. A computational GRID is typically heterogeneous in the sense that it combines clusters of varying sizes, and different clusters typically contains processing elements with different level of performance. In this, heuristic approaches based on particle swarm optimization and ant colony optimization algorithms are adopted for solving task scheduling problems in grid environment. Particle Swarm Optimization (PSO) is one of the latest evolutionary optimization techniques by nature. It has the better ability of global searching and has been successfully applied to many areas such as, neural network training etc. Due to the linear decreasing of inertia weight in PSO the convergence rate becomes faster, which leads to the minimal makespan time when used for scheduling. To make the convergence rate faster, the PSO algorithm is improved by modifying the inertia parameter, such that it produces better performance and gives an optimized result. The ACO algorithm is improved by modifying the pheromone updating rule. ACO algorithm is hybridized with PSO algorithm for efficient result and better convergence in PSO algorithm.

Keywords

Pheromone, Swarm Intelligence, Inertia, Grid Scheduling.

Full Text