A hybrid optimization algorithm for energy-aware multi-objective task scheduling in heterogeneous multiprocessor systems

In order to successfully address the multiprocessor work scheduling issues, a novel hybrid approach is presented in this paper. Executing a group of tasks on a small number of processors is the main goal of a task scheduling method. The job scheduling issue is viewed as a multi-objective optimisation issue in this study. The goals of this optimisation challenge are to determine the system’s least energy consumption and minimum makespan (the sum of the schedule lengths for each task’s execution). In order to lower the processors’ energy usage, the Dynamic Voltage Frequency Scaling (DVFS) level has the task scheduling issue enabled. To determine the goals, a hybrid algorithm called Hybrid Grey Wolf Crow Search Optimisation (HGWCSO) is suggested. The Grey Wolf Optimisation (GWO) and Crow Search Optimisation (CSO) algorithms are combined to create the HGWCSO algorithm. The social structure and hunting strategy of wolves serve as the inspiration for the GWO metaheuristic algorithm. The CSO, on the other hand, is a metaheuristic algorithm that was motivated by crows’ clever conduct. The proposed hybrid approach is put into practise using tasks that are produced at random, and the results of this comparison are classic methods. The proposed method’s simulation results are contrasted with those from five existing widely used methods. The proposed technique has also been contrasted with some newly created metaheuristic algorithms using various real-world data sets. According to the simulation results, HGWCSO outperforms other algorithms in terms of lowering makespan and energy usage. The proposed algorithm has successfully improved the makespan and energy consumption with a maximum of 31.07% and 58.36%, respectively, compared to other algorithms.