Economic-based ACO Algorithm for Data Intensive Grid Scheduling

• Published in: Asian journal of scientific research Vol. 6; no. 4; p. 789
• Main Authors: Aranganathan, S, Mehata, K M
• Format: Journal Article
• Language: English
• Published: 2013
• Subjects: Formicidae
• ISSN: 1992-1454; 2077-2076; 2077-2076
• DOI: 10.3923/ajsr.2013.789.796

The scope of grid computing is rapidly growing in distributed heterogeneous environments for the need to utilize and share large-scale resources to solve complex scientific problems. Economic models are effective in collaborating large-scale heterogeneous data and computational resources that are typically owned by different organizations with diverse interests. Scheduling is the most crucial task to achieve high performance in both computation and data grids. To utilize the grid efficiently for both resource providers and consumers, an efficient job scheduling algorithm is required. The proposed algorithm allows resource providers and consumers to take autonomous scheduling decisions and that both parties can derive sufficient incentives based on their economic interests. It is based on the general adaptive scheduling heuristic which employs a Quality of Service (QoS) guided component that emphasizes more on reliability. The algorithm was successfully tested in simulation environment. Experiments showed that the proposed economic and ant heuristic method was able to significantly improve performance by 10-25% even in unreliable network conditions.