Improved Black Widow Optimization: An investigation into enhancing cloud task scheduling efficiency

The Black Widow Optimization (BWO) algorithm has garnered significant attention within the realm of metaheuristic algorithms due to its potential to address diverse problems across various domains. However, a noteworthy weakness of BWO is its utilization of a random selection technique, which can lead to reduced diversity, expedited convergence, and potential entrapment in local optima. This research introduces a novel approach to augment the BWO algorithm by integrating alternative selection schemes, thereby surpassing the limitations of the current selection methodology. To assess the effectiveness of these proposed variants, we employ the CEC 2019 benchmark functions as the standard evaluation metric. Subsequently, we utilize the best-performing BWO version, PIBWO, to address cloud scheduling challenges. In a series of comparative experiments, PIBWO demonstrates superior performance compared to existing algorithms, showcasing remarkable enhancements in makespan reduction, energy consumption minimization, and cost efficiency. These findings underscore PIBWO's potential as a robust solution for addressing cloud task scheduling challenges, offering promising avenues for developing more sustainable and cost-effective cloud computing systems.