A task allocation method for heterogeneous multi-core system based on genetic algorithm

Abstract

Heterogeneous multi-core platforms are increasingly prevalent due to perceived superior performance over homogeneous systems. In order to maximize performance, each task needs to be mapped to the most appropriate processor. This paper implements a task allocation method based on genetic algorithm. The genetic algorithm is used to sample the application load feature in the task scheduling time slice, and its complicated iterative process is distributed to the following multiple scheduling sampling periods to select the core which complies with its calculation characteristic for each task. Experimental results demonstrate that the algorithm can effectively improve the system performance, compared with the built-in task scheduling mechanism of Linux 2.6 kernel.