STDO: Siberian Tiger and Devil Optimization — A Novel Hybrid Metaheuristic Algorithm for Energy-Efficient Task Scheduling in Cloud Computing

Abstract

Energy-efficient task scheduling has emerged as a critical research challenge in cloud computing due to the exponential growth of data centres and associated energy consumption. Existing metaheuristic algorithms such as Particle Swarm Optimization (PSO), Tasmanian Devil Optimization (TDO), and Siberian Tiger Optimization (STO) suffer from limitations including premature convergence, excessive exploration, and stagnation in complex search spaces. This study proposes a hybrid algorithm, Siberian Tiger and Devil Optimization (STDO), which integrates exploration and exploitation mechanisms through a phased switching strategy and a persistent elite archive. The proposed method is evaluated across twelve heterogeneous cloud configurations with varying virtual machine capacities and task loads. Each experiment is conducted over multiple independent runs to ensure statistical reliability. The results demonstrate that STDO achieves superior energy efficiency compared to baseline algorithms while maintaining competitive makespan performance. Statistical validation using the Wilcoxon signed-rank test confirms the significance of improvements. The findings establish that hybrid metaheuristic approaches can effectively enhance scheduling performance in cloud environments while ensuring scalability and robustness. STDO was evaluated in comparison to TDO, STO, and PSO in 12 cloud configurations. These configurations included three VM pool sizes (5, 10, and 20 VMs) and four task levels (50 to 200 tasks). With an average of 0.024456 MI/Watts, STDO outperformed STO and PSO in every configuration, outperforming TDO by 12.0%, STO by 6.0%, and PSO by 23.9% in terms of energy efficiency. On scheduling time, it reduced makespan by up to 33.47% over PSO and 30.79% over TDO in constrained settings. It also held up far better as task loads scaled up, where PSO degraded by as much as 29% and STDO remained comparatively stable.