Workflow load balancing using soft computing base novel framework with Qos parameters

Abstract

Cloud computing represents a new era of computing network, where the resources of the system are dispersed and shared among its users in the network premises. The user of this system is able to use such resources through the technology of internet based on system of Pay-As-Per-Use. If a service is used by any type of user, it helps in production of wide variety of data. So, the cost of data transfer between two of the dependent resources will be extremely high. Additionally, an application of complex nature involves large number of tasks boosting the process of total cost of execution with respect to the used application, if the process is not scheduled in an optimized manner. In order to overcome such issues, a hybrid approach of water cycle optimization is proposed with particle swarm optimization. This method is divided into two steps of working determining under and over utilized virtual machines. In experimental analysis, the proposed approach on different scientific workflows is done where significant performance in all the workflows is based on total execution time and total execution cost.