Enhanced Task Scheduling With Metaheuristics for Delay and Energy Optimization in Cloud-Fog Computing

The growth of the Internet of Things (IoT) and its application across various industries has produced large volumes of data for processing. Tasks that require prompt responses, particularly delay-sensitive ones, are directed to the nearest fog nodes. Offloading critical tasks to the cloud reduces user-side energy consumption but increases latency due to longer transmission distances. Fog nodes, being closer to the source, minimize delay but may require more local energy. Another major issue in cloud-fog computing is allocating tasks to suitable resources according to task needs. To tackle these challenges, this study introduces a hybrid meta-heuristic approach by combining the Butterfly Swarm Optimization (BSO) algorithm with the heuristic Minimum Completion Time (MCT) initialization method. The key innovation of this work lies in the integration of MCT-based heuristic initialization with the BSO algorithm, enabling faster convergence and more efficient task scheduling by balancing energy and delay in heterogeneous cloud-fog environments. Both delay and energy consumption are reduced through the MCT-BSO algorithm, in which the fog broker effectively manages the task distribution. Simulation results show that the MCT-BSO method achieves delay reductions of approximately 20.7% to 36.3% and improvements in energy consumption ranging from 15.4% to 38.1%, significantly outperforming comparative algorithms such as Grey Wolf Optimization, Nondominated Sorting Genetic Algorithm II, and Modified Particle Swarm Optimization, particularly under high workload conditions.