Task scheduling in cloud computing systems using multi-objective honey badger algorithm with two hybrid elite frameworks and circular segmentation screening

Abstract

In cloud computing environment, task scheduling is the most critical problem to be solved. Two different multi-objective honey badger algorithms (MOHBA-I and MOHBA-II) based on hybrid elitist framework and circular segmentation screening are proposed for the multi-objective problem of task scheduling optimization in cloud computing systems. MOHBA-I and MOHBA-II combine the grid indexing mechanism and decomposition technique, respectively, to select better populations based on elite non-dominated sorting. A circular segmentation screening mechanism was proposed to retain the superior individuals when the regional density is too high to further maintain the diversity of the populations, and attach an external archive to preserve the uniformly diversified Pareto decomposition set. The performance of the proposed algorithms is verified by using test functions. MOHBA-I and MOHBA-II achieve the first and third rankings, respectively, compared to other classical multi-objective algorithms. Solve the cloud computing task scheduling problem using time, load and price cost as metrics, test for different task sizes, and compare MOHBA-I with algorithms such as NSGA-III, MOPSO and MOEA/D in the same experimental environment. When facing a large-scale task, MOHBA-I ranks first in HyperVolume value with 2.4449E−02 for two objectives and 9.2950E−03 for three objectives. The experimental results show that MOHBA-I finds the highest number of solutions with better convergence and coverage, obtaining a satisfactory Pareto front, which can provide more and better choices for decision makers.