Adaptive Random Evolution Whale Optimization Algorithm (AREWOA) Based Workflow Scheduling in a Virtualized Cloud Environment

Cloud workflow task scheduling refers to assigning workflow tasks submitted by users to appropriate computer resources for execution. The workflow scheduler assigns the user request to the available Virtual Machine (VM) hosted on cloud servers during scheduling. However, while limiting energy consumption, the workflow application processing schedule length (makespan) and associated costs must stay within the users budget and deadline. The tasks are chosen using the optimal scheduling technique to reduce Energy Consumption (EC) and maximize Resource Utilization (RU) without compromising system performance. In this paper, Adaptive Random Evolution Whale Optimization Algorithm (AREWOA) based scheduling is proposed to reduce EC, maximize RU, and workflow execution time. AREWOA is performed based on the dual-operation strategy partnership. Initially, the Gaussian distribution is used in the randomization procedure to raise the variety of the population. Secondly, a new skip step factor is introduced to enhance the optimizer capability to solve local optimum, and specific reinforcement is applied to the hunter process to enhance the whale’s exclusive exploration or exploitation skills. An adaptive weight factor is applied to maintain algorithm stability and a balance among exploration and exploitation. Dynamic provisioning of infinite heterogeneous resources, AREWOA is thought to demonstrate cloud-based Infrastructure-as-a-Service (IaaS). This approach is simulated in Cloud Simulator and the results shows that the efficiency of the methods.