A hybrid evolutionary framework for efficient IoT task scheduling in fog computing

This paper proposes an innovative, hybrid approach to task scheduling and virtual machine (VM) placement in fog computing environments that aims to optimize energy efficiency and task completion for Internet of Things (IoT) applications. The proposed method combines Gorilla Troops Optimizer (GTO), a bio-inspired metaheuristic, with a resource-aware virtual machine placement strategy that allows for simultaneous optimization of task scheduling and virtual machine allocation. This integrated approach addresses key challenges in the allocation of IoT tasks by considering multiple objectives, such as latency, power consumption, and load balancing. A dynamic exploration-exploitation strategy and innovative fitness functionality have been employed to efficiently map tasks to fog nodes while minimizing task failure and suspension periods. Extensive simulations performed with iFogSim2 demonstrate the effectiveness of the proposed method and have achieved significant improvements over existing algorithms. The proposed approach is 18% better than ant colony optimization (ACO), 15% better than improved multi-objective differential evolution (IMODE), and 13% better than genetic and simulated annealing (GASA). These results underscore the effectiveness of the hybrid method in optimizing task scheduling and VM placement for dynamic and latency-sensitive applications in IoT. This work provides a scalable solution for fog computing systems that significantly improves service quality by optimizing resource consumption and reducing energy consumption, providing a promising approach to real-world IoT environments.