Energy Optimization for DVFS-enabled CPS through Reliable and Real-time Task Mapping

Improved energy efficiency, real-time response, and execution reliability are important objectives during the task mapping process in Cyber-Physical Systems (CPS). Our goal is to develop an effective task mapping scheme to balance these conflicting objectives. To achieve this goal, dependent tasks are duplicated to enhance the reliability and mapped on the wireless nodes of CPS with Dynamic Voltage/Frequency Scaling (DVFS) capabilities to reduce the energy consumption, while still meeting the real-time constraints. Our contributions include the formulation of task mapping problem as an Integer Non-Linear Programming (INLP) to balance the energy consumption of the nodes with real-time and reliable task execution and the corresponding problem linearization. In addition, we propose a novel heuristic method to solve this task mapping problem. The performance of the proposed method is evaluated through a series of simulations.