Communication energy oriented scheduling optimization for network on chip

It is difficult to improve the performance for the traditional processor architecture. The power consumption and the temperature of the processors are the main bottlenecks for CPU with single core. Chip multiprocessor is proposed to solve such problems. In order to solve the communication problem between the on chip components, network on chip is taken as the promising diagram. The on-chip cores are connected by lines. However, the communication is now the core of the on-chip network and the power consumed is more than before. This paper aims at the energy consumption problem in communication. The tasks will be mapped to the cores according to the communication density to reduce the power consumption and improve the performance. The experimental results show that our algorithm can achieve its target.