Business Process Driven Trust-Based Task Scheduling

Abstract

The demand for agile and flexible business application systems has sparked interest in using cloud computing technology to respond quickly and effectively to a dynamic business environment. The authors classify the appropriate cloud services as a multi-objectives task scheduling problem in a hybrid cloud service system. In this article, the authors propose a business process (BP) driven task scheduling system that supports multiple clouds, including private ones. A trust-based non-dominated sorting genetic algorithm (NSGA2) is developed to solve the multi-objective task scheduling problem. By sorting populations into different hierarchies based on the ordering of Pareto dominance, they identify a Pareto-optimal multi-dimensional frontier that permits managers to reconcile conflicting objectives when scheduling tasks on cloud resources. The authors illustrate the usability and effectiveness of their approach by applying it to a case study conducting simulated experiments.