ABSTRACT
Power and related metrics have become important design constraints in computing systems ranging from low-end embedded systems to high-end servers for various reasons. The limited improvements in the energy capacity of batteries and the rapid growth in the complexity of battery-operated systems have reduced the longevity of operation between battery recharges. Hence, optimizing energy consumption is of crucial importance in battery driven mobile and embedded devices. Energy optimization is specifically important in sensor nodes that may not be accessible after deployment for replacing the batteries. Hence, energy optimization is critical in determining the longevity of the sensor network. Due to constraints on physical size/weight, the energy capacity of the battery in sensor nodes is typically limited. Hence, energy optimization can either increase the lifetime of the node or enable more powerful operations at a sensor node. In order to tackle the limited battery capacities, many sensors scavenge energy from the external environment in addition to using batteries. However, the amount of power that can be harnessed this way is very limited (typically, in the range of 100 μW). Therefore, energy optimizations are important even in these environments. Energy consumption has a significant impact on the weight 596of the battery pack and consequently of the entire sensor node. For example, to limit the re-charge interval to 10 h, it requires a 0.5 lb battery to operate a system that consumes 1 W. *
