Cost Efficient Genetic Algorithm for Resource Scheduling in Cloud Infrastructure

Abstract

Cloud computing is a reliable computing platform for large computational intensive or data intensive tasks. This has been accepted by many industrial giants of software industry for their software solutions, companies like Microsoft, Accenture, Ericson etc has adopted cloud computing as their first choice for cheap and reliable computing. But which increase in number of clients adopting this there is requirement of much more cost efficient and high performance computing for more trust and reliability among the client and the service provide to guarantee cheap and more efficient solutions. So the tasks in cloud need to be allocated in an efficient manner to provide high resource utilization and least execution time for high performance, at the same time provide least computational cost. Many resource algorithms are been proposed to improve the performance, but are not cost efficient at same time. Algorithms like genetic, particle swarm and ant colony algorithm are efficient solutions but not cost efficient. So to overcome these issues, we have proposed a learning based cost efficient algorithm for cloud Infrastructure. Proposed algorithm uses genetic algorithm for cost efficient task allocation to minimize cost and high utilization to provide better QoS (Quality of Service) to the client. Proposed strategy has proven to have better performance in term of execution cost, execution time, scheduling time as compared to previously proposed task allocation algorithm