An Integrated Approach for Scheduling Divisible Load on Large Scale Data Grids

Abstract

In many data grid applications, data can be decomposed into multiple independent sub datasets and distributed for parallel execution and analysis. This property has been successfully exploited for scheduling divisible load on large scale data grids by Genetic Algorithm (GA). However, the main disadvantages of this approach are its large choromosome length and execution time required. In this paper, we concentrated on developing an Adaptive GA (AGA) approach by improving the choromosome representation and the initial population. A new chromosome representation scheme that reduces the chromosome length is proposed. The main idea of AGA approach is to integrate an Adaptive Divisible Load Theory (ADLT) model in GA to generate a good initial population in a minimal time. Experimental results show that the proposed AGA approach obtains better performance than Standard GA (SGA) approach in both total completion time and execution time required.