Energy Efficient Task Allocation and Scheduling in Distributed Homogeneous Multiprocessor Systems

With the advent of semi conductor technology, the development of more complex embedded real time applications is made possible today. This accelerates the development and support for multiprocessor based systems. The paper presents the development of “a power-aware real time embedded system for temperature monitoring and control in safety critical applications”. The main objective of the work is to perform a hardware implementation of task allocation for multiprocessor systems based on task dependencies and precedence relations and schedule the real time tasks. The proposed work also provides an energy-aware solution by integrating Dynamic Voltage Frequency Scaling (DVFS) technique and shutting off the unused peripherals of the processor. A laboratory model of multiprocessor based experimental setup was developed to validate the functioning of algorithms using ARM7 LPC2148 microcontrollers. Key-Words: Multiprocessor systems, Energy minimization, Embedded Systems, Microcontrollers, Scheduling, Task-processor allocation, Dynamic voltage frequency scaling