Convergence of the Harris hawks optimization algorithm and fuzzy system for cloud-based task scheduling enhancement

Task scheduling entails the allocation of various tasks to virtual machines. Consequently, scheduling algorithms are meticulously crafted to achieve an array of objectives, including the reduction of makespan, the minimization of energy consumption, the enhancement of resource productivity, the attainment of load balancing, and the optimization of costs. Given the profound importance of these goals, algorithms tailored for such scenarios invariably encompass multiple objectives. This research paper introduces an innovative multi-objective task scheduling algorithm for cloud computing, which seamlessly integrates the Harris hawks optimization (HHO) algorithm and incorporates the power of fuzzy logic. Dubbed the "fuzzy-HHO" methodology, it harnesses the HHO algorithm to explore the expansive solution space while subjecting the generated solutions to meticulous evaluation through fuzzy logic. The HHO algorithm unfolds in two distinct phases: exploration and exploitation. Within the exploitation phase, a cascade of four stages is executed: soft besiege, hard besiege, soft besiege with progressive rapid dives, and hard besiege with progressive rapid dives. This intricate algorithm offers robust strategies to effectively navigate away from local optima, rendering it proficient at approximating and even converging upon global optima. To substantiate its efficacy, the proposed method is rigorously compared against two state-of-the-art algorithms within the CloudSim framework. Through meticulously conducted simulations, compelling evidence emerges, the proposed method consistently outperforms the comparison algorithm by remarkable margins-up to 47% enhancement in makespan reduction, 73% decrease in energy consumption, and an impressive 19% cost reduction. These substantial improvements are particularly evident in scenarios encompassing a substantial number of tasks (10,000 tasks).