Optimizing the energy efficiency of embedded systems, using techniques such as instruction level measurements and software/hardware component simulations, focus their ability to monitor and optimize an embedded system within the confines of the laboratory. These methods consume a large amount of time and money since the designer must continually reprogram the embedded system and have the capability of measuring currents ranging from micro amps to amps within fractions of a second. Unfortunately, optimizations made in the lab can easily be undone in the environment the device is deployed in as a result of temperature changes and variations in externally connected hardware (radios, memory modules etc.). The proposed energy efficiency optimization method can monitor the energy consumption of an embedded system in real-time, and perform hardware configuration optimizations after the device has been in operation. |