Prediction of Task Execution Time in Cloud Computing

Abstract

Now-a-days all cloud service providers are providing all resources to end users at very cheap rate and at the same time by designing scheduling algorithms cloud service providers are ensuring that all users can get response data in quick time. Proper utilization of resources can be enhanced by better resource allocation which will decrease the total makespan and the response time of the task. In this proposed work, a novel scheduling algorithm called Prediction of Execution Time of Tasks algorithm (PETT) to estimate minimum task execution time for cloud computing environment is proposed. In PETT algorithm, an array with all tasks and processors is built as a Directed Acyclic Graph (DAG). The matrix contains estimated execution task time on each processor and all rows of the matrix are filled with all processors execution time for all tasks. On generated matrix, PCA algorithm has been applied to choose that processor, which consumes less execution time for selected task. This process continues till all tasks assigned to all processors. By applying PETT algorithm, further computation and communication cost can be reduced at the cloud end.