Performance Analysis of Computational Grid Job Scheduling using Bio-Inspired Heuristic Function

Abstract

The increasing rate of jobs and a limited number of resources decline the performance of the computational grid. Therefore, the scheduling of tasks plays a vital role in the computational grid. The conventional scheduling of the computational grid applies the CPU scheduling algorithms such as FCFS, SJF and round-robin. However, the limited constraints factors of scheduling algorithms increase the ratio of job failure and degrade the overall performance of computational grids. Therefore,
the incremental research approach uses a bio-inspired heuristic function to focus on the task scheduling algorithm. Furthermore, the searching capacity of the bio-inspired function increases the utilization of resources such as CPU and memory to share resources. This paper presents the experimental analysis of various algorithms such as ACO, PSO, GSO, ABC and TLBO to schedule tasks in a computational grid of different sizes such as 10 X 10, 20X20 and 40 x 40. the simulation software uses MATLAB version R2014a. The empirical evaluation of the systems is estimated with job failure and job completion.