Comparitive Analysis of ACO and PSO in Grid Job Scheduling

Abstract

Grid computing is a one that coordinates and shares computation, application, data storage, or network resources across dynamic and geographically dispersed organizations. Scheduling is one of the main critical design issues of grid computing. It becomes a challenge because the capability and availability of resources vary in a dynamic nature. The complexity of scheduling problem increases with the size of the grid and becomes difficult to solve effectively. To solve this issue we go for designing new optimal solutions. It mainly focuses on new heuristic techniques that provide an optimal or near optimal solution for larger computational grids. Ant Colony Optimization based scheduling algorithm and PSO the population based search algorithm is used .The PSO is mainly based on the simulation of the social behavior of bird flocking fish schooling approach. The proposed scheduler allocates an application to a host from a pool of available hosts and applications by selecting the best match. Based on the experimental results, we prove that the proposed PSO algorithm confidently demonstrates its practicability and competitiveness with ACO algorithms.