Token based energy aware scheduling algorithms for heterogeneous multi-core

Abstract

In this paper, Energy aware scheduling algorithms are developed for task allocation on heterogeneous multi-core processors is NP-hard. The objective is to create efficient scheduling method to solve problem. ETPS (Energy token proportionate sharing) algorithm and dynamic ETPS are the scheduling methods to generate optimality for task mapping on multi core. To validate the model extensive simulation were carried out in C++ simulator. MILP formulations are developed to create token based scheduling algorithms. Our simulated results achieve better performance and reduced 15% of task miss rate and maximize the energy efficiency.