Optimized Task Offloading in Hybrid Cloud-Fog Computing for IoT Using Horse Herd Optimization Algorithm (HHOA)

One of the main obstacles to enhancing the functionality is task offloading for Internet of Things (IoT) applications in hybrid cloud and fog computing environments. This study looks into an effective task offloading strategy that uses the Horse Herd Optimization Algorithm (HHOA) to reduce execution time and resource costs while ensuring balanced resource use. Unlike most previous studies that depend on synthetic datasets, this work uses a real-world IoT fog and cloud dataset sourced from Kaggle, which makes the evaluation more realistic and relevant. The dataset includes 120 tasks spread across 13 computing nodes, featuring 10 fog nodes and 3 cloud nodes, each with different processing capacities and resource costs. Simulation results show that HHOA achieves a total execution time of 0.9731 seconds, an optimal scheduling cost of 2.155, and a makespan of 6.1296 seconds. The highest recorded execution time per task was 0.8757 seconds. The highest execution cost per task was 0.8712. This shows minimal variation and balanced load distribution. The algorithm smartly assigned time-sensitive tasks to cloud nodes for quicker processing. It allocated less urgent tasks to fog nodes to preserve their limited CPU, RAM, and power resources. These findings confirm that HHOA effectively delivers a cost-efficient, performance-optimized, and resource-conscious method of job offloading for cloud computing and hybrid IoT fog settings.