DLH-GWO: a dimension learning and hunting-enhanced grey wolf optimizer for multi-objective internet of things scheduling in fog environments

Effective task scheduling is critical to minimizing energy consumption and optimizing resource utilization in Internet of Things (IoT) applications deployed on fog computing platforms. This paper proposes a novel algorithm, Dimension Learning and Hunting-optimized Grey Wolf Optimization (DLH-GWO), to optimize green IoT scheduling in fog computing. DLH-GWO combines the basic GWO with a dimension-learning scheme and a local-hunting scheme to improve the exploration–exploitation balance and accelerate convergence. It optimizes solution diversity and avoids early convergence, dramatically cutting energy consumption, task makespan, and rejection rates in fog-based IoT systems. From the experimental results, DLH-GWO is observed to significantly outperform other metaheuristics in terms of energy consumption and makespan. The DLH-GWO algorithm has been shown to perform efficiently regardless of the number of tasks and the capacities of fog nodes. This paper develops a highly efficient optimality method for the emerging field of fog computing and a scalable solution to IoT scheduling problems.