Efficient Energy Aware Task Scheduling In Heterogeneous Network On Chip

Document Type : Primary Research paper

Authors

1 Assistant Professor, Department of Computer Science and Engineering, Bannari Amman Institute of Technology

2 Professor, Department of Electrical and Electronics Engineering, Kumaraguru College of Technology, Coimbatore

3 Assistant Professor, Department of Information Technology, Bannari Amman Institute of Technology

Abstract

High performance, Quality of Service and energy efficiency are the main factors
in real-time embedded applications. Network-on-Chip (NoC) based Multiprocessor Systemon-
Chips (MPSoC) computing platform performs better than superscalar uniprocessor
architectures. The operating cost, lifetime and readability of the embedded system are
based on the energy consumption. Reduction of energy consumption can be accomplished
by efficiently scheduling the tasks with parameters such as priority and constraints.
Ordering of tasks, assignment of voltage for task mapping need to be considered in Energy
aware scheduling. Contention between the communications plays a major role in Energy
consumption. Ordering of tasks, tasks mapping, assignment of voltage and contention
should be combined for energy optimization. Energy consumption can be reduced by
mapping each task to processor and minimizing the energy consumed in communication
by allocating distinct levels of voltage to the Links in NoC. Scaling of frequency, dynamic
voltage and power management can be used for optimization of energy consumption. The
proposed approach maps contention-aware tasks and manages the energy by ordering the
task and assigning the priority to the tasks.

Keywords