A Review of Techniques for Power Optimization at Different Levels of Abstraction of Low Power Embedded Systems

ABSTRACT The markets for “Anywhere-Anytime” computing devices and the Internet of Things (IoT) are being established rapidly as advanced versions of these devices have become available to the consumers. The proliferation of these intelligent portable computing devices has created an increasing demand for low power embedded systems, which has resulted in the fabrication of tiny high-speed processors with higher packaging density and powerful computing capabilities. These portable devices are expected not only to be small, lightweight and cool but also to provide incredibly long battery life. This requirement has presented significant challenges for managing the power in low power embedded systems, and it has shifted the designers' focus from traditional constraints to power consumption. This paper reviews the various power optimization techniques that can be applied at different levels of abstraction to provide an in-depth view into the working of power optimization techniques for low power embedded systems.