IGWOA: Improved Grey Wolf optimization algorithm for resource scheduling in cloud-fog environment for delay-sensitive applications

Fog computing, a technology that offers adaptable and scalable computing resources, facing a significant difficulty in task scheduling, affecting system performance and customer satisfaction. Finding solutions to the task scheduling problem is challenging due to its NP-completeness. Researchers suggest a hybrid approach that combines the Grey Wolf Optimization Algorithm (GWO) and Heterogeneous earliest finishing time (HEFT) to address this problem. The hybrid IGWOA (Improved Grey Wolf optimization algorithm) method seeks to minimize makespan and throughput while focusing on multi-objective resource scheduling in Fog computing. Proposed algorithm is suggested to improve the exploration and exploitation phases of the traditional grey wolf algorithm. Furthermore, the HEFT-based GWO algorithm has the benefit of faster convergence in larger scheduling problems. The effectiveness of the suggested algorithm in comparison to existing techniques has been evaluated using the iFogsim toolkit. Real data set and pseudo workloads both are used for working. The statistical method Analysis of Variance (ANOVA) is used to confirm the results. The effectiveness of it in reducing makespan, and throughput is demonstrated by experimental results on 200–1000 tasks. Particularly, the proposed approach outperforms peer competing techniques AEOSSA, HHO, PSO, and FA in relation to makespan and throughput; successfully, improvement is noticed on makespan up to 9.34% over the AEOSSA and up to 72.56% over other optimization techniques for pseudo workload. Additionally, it also showed improvement on makespan up to 6.89% over the AEOSSA and up to 69.73% over other optimization techniques on NASA iPSC and HPC2N real data sets, while improving throughput by 62.4%, 52.8%, and 41.6% on pseudo workload, NASA iPSC, and HPC2N data sets, respectively. These results show proposed approach solves the resource scheduling issue in Fog computing settings.