Evaluating the Performance of Various MOEA's to Optimize Scheduling Overhead in Homogeneous Multicore Architecture
R. Lavanya, S Sivarani, Cecil Prabhaker. M Lordwin, T Jeyalakshmi, M Muthulakshmi · 2018 International Conference on Current Trends towards Converging Technologies (ICCTCT) · 2018
In multi core processors, the considerable factor that determines the performance is task scheduling. In order to optimize scheduling overhead in such processors, the major multiple objectives to be considered are maximized speed, minimized temperature and power consumed during the process. The four multi objective scheduling algorithms used here are MOGA, NPGA, NSGA and PAES. In this work, the evaluation and the study for which of these scheduling strategies achieves better performance is taken. The overall performance is estimated by calculating power consumption, core utilization and temperature. In this evaluation, PAES is regarded as the best one since it alleviates the drawbacks of other three algorithms.