Bradley Terry Brownboost and Lemke flower pollinated resource efficient task scheduling in cloud computing

Abstract

Cloud computing (CC) is extensively used across various domains, yet task and resource scheduling still demand significant improvement. In heterogeneous computing systems, effective task scheduling ensures optimal task-machine mapping, reducing makespan and enhancing resource utilization. One major challenge in cloud data centers is managing vast user requests while maintaining efficient scheduling. This work introduces the Bradley–Terry BrownBoost and Lemke flower pollinated resource optimization (BTB-LFPRO) method to enhance task scheduling and improve performance. The BTB-LFPRO approach includes two main steps: classification and optimization. First, the Bradley–Terry BrownBoost Classifier categorizes tasks into high- and low-priority based on pairwise comparisons. Then, the Lemke flower pollinated resource optimization algorithm selects the optimal virtual machine using swarm intelligence. This algorithm balances global exploration and local exploitation via Lévy flights to find the best scheduling path. Experimental results demonstrate that the BTB-LFPRO method significantly improves task scheduling efficiency by 24% and enhances throughput by 24%, outperforming existing techniques.