A Memetic Genetic Algorithm for Optimal IoT Workflow Scheduling

Amer T. Saeed, Gang Chen, Hui Ma, Qiang Fu

Abstract: Internet of Things (IoT) devices have become a crucial part of daily life. Because IoT devices often have small processing capability and low power supply, two popular technologies, i.e. cloud servers and fog edges, are increasingly integrated with IoT for workflow execution, giving rise to the resource allocation and workflow scheduling problem in hybrid IoT environments, i.e. the IoT workflow scheduling (IoTWS) problem. To tackle this NP-hard IoTWS problem, a new Genetic Algorithm (GA) called IoTGA has been successfully developed in this paper. In comparison to state-of-the-art GA approaches from literature, IoTGA allows fast workflow execution and can explicitly reduce the time and energy consumption thanks to its use of a newly designed local search method. Experiments on benchmark IoTWS problems clearly indicate that IoTGA can significantly outperform several competing GA methods and are more useful in practice.