An efficient astronomical image processing technique using advance dynamic workflow scheduler in cloud environment

Abstract

The Montage project of NASA introduces an efficient astronomical image mosaic processing service with the help of Montage workflow applications. These workflow applications contain a set of input astronomical images that are able to produce custom mosaics of the sky to understand the universe’s activities. In the last 2 decades, extensive research has been done for processing Montage application in the various distributed computing paradigms such as Grids and Clusters but fails to achieve the required performance in terms of makespan. Recently, Cloud computing is the latest computing paradigm that provides online computing resources for efficiently processing these applications. In this article, we introduce an efficient astronomical image processing technique (EAIPT) for Montage workflow application in the cloud environment that shows notable improvements in terms of minimization of makespan without increasing the time complexity with the help of an advance dynamic scheduler. The proposed dynamic scheduler tries to schedule the current workflow task on the virtual machine that already processed its last parent task based on proposed scheduling constraints. In this way, it bypasses the communication time among different VMs. The performance of the proposed algorithm is compared with the existing state of art algorithms. The experimental results prove that our proposed technique has significant performance over compared algorithms in terms of schedule length ratio (SLR), the percentage of best results, and average running time metrics.