Real-time embedded systems (RTES) found in many real-world applications demand that timing requirements for computations and communications are satisfied. In addition, energy consumption is becoming a major consideration for such systems, because of the proliferation of mobile, wireless, and embedded systems with limited energy resources. Energy consumption significantly impacts the cost, performance, and life time of such systems. This research addresses the lack of system-wide energy conservation approaches by developing integrated scheduling techniques to reduce the overall energy consumption of wireless real-time systems while meeting applications' timeliness requirements. Integration in this context refers to reducing the energy consumption of a system as a whole by maximally exploiting the Dynamic Voltage Scaling (DVS) capabilities of CPU cores and the Dynamic Power Management (DPM) capabilities of other resources and devices. Important results of this work are novel and highly applicable algorithmic techniques for energy-efficient RTES. The broader impacts of this project can be found in both education and research, in addition to its enormous social impacts due to the pervasiveness of real-time systems and environmental concerns. The results of the experimental studies are being made available online for other researchers to pursue their own research and education interests. The effort also helps greatly to alleviate the problem of many theoretical system-level power reduction techniques not being adopted due to the lack of sound studies based on actual hardware platforms.
