Critical Path-Based Ant Colony Optimization for Scientific Workflow Scheduling in Cloud Computing Under Deadline Constraint

In local computing environment, when we are dealing scientific computation using scientific workflow scheduling environment under deadline constraint, QoS is one of the most challenging tasks for any system used in scientific computing systems. Because when we are focusing on minimizing the workflow execution cost as well as time, we should not forget to consider the user-defined quality of service requirements while minimizing the workflow execution of cost and time. Therefore, to reduce the cost and time, we used cloud environment. Since cloud computing environment is in elastic nature, in which availability of resources is readily available as when and then required, its utilization is another challenge while using cloud computing environment. Therefore, in this paper, we use intelligence ant colony optimization (ACO), in which underutilized VMs allocation is initialized by Pareto distribution. ACO is used to converge the decision of virtual machine (VM) migration by its convergence to minima of cost and time. In our experiments, we have set up a local simulator in stand-alone system; for that workflow simulator, 1.0 has been used, where we have used Java eclipse for executing our program to calculate total execution time (TET) and total execution cost (TEC). In which, we have run our simulator ten times for each scientific workflow application and then average is taken for each workflow application to compare the output in which we found that ACO shows significant performance component when compare to existing genetic algorithm.