Energy-Efficient Task Scheduling in Fog-Cloud Environments Using an Improved Single Candidate Optimizer

Abstract

Fog-cloud computing has emerged as a promising paradigm for supporting latency-sensitive and data-intensive Internet of Things (IoT) applications through the use of fog computing to bring cloud-style functions closer to the end user. Task management in these heterogeneous environments is challenging due to dynamic workloads and resource constraints; thus, there are also difficulties in balancing the trade-offs among energy consumption, execution time, and operational cost. Metaheuristic-based scheduling algorithms have typically exhibited relatively high computational complexity, slow convergence rates, and poor adaptability to real-time conditions. In this study we present an energy efficient task scheduling framework using an improved single-candidate optimizer (ISC-F). The proposed algorithm improves the existing single-candidate optimizer (SCO) by using a two-phase optimization method that combines exploration and exploitation effectively. In addition, ISC-F allows fog nodes to be prioritized for the execution of latency-sensitive tasks, while dynamically offloading tasks to cloud resources as network conditions change. The optimization process considers multiple objectives including minimum energy consumption, makespan, execution time, and total cost. To evaluate the potential of ISC-F against competing scheduling algorithms, we conducted a series of extensive simulations comparing the aforementioned competing scheduling methods (HEFT, IHEFT, IKH-EFT, IWO-CA). The results demonstrate that ISC-F significantly outperforms existing scheduling techniques; therefore, there is strong evidence that the proposed optimization algorithm provides significant benefits to both fog-cloud computing systems and their respective users.