Sustainable and efficient energy management for mobile and networked systems

Mobile systems (such as smart phones) and networked systems (such as wireless sensor networks) have evolved into a key technology for applications including health monitoring, intelligent transportation, and infrastructure protection. To support these long-term applications, energy management is extremely important. This is because smart phones and sensor nodes are normally equipped with limited amount of batteries to reduce size and cost, while battery technology is not evolving at the same pace as applications. Slow development in battery technology and rapid advances in ultra-capacitor design have motivated us to investigate the possibility of using capacitors as the sole energy storage units for mobile and networked systems. Furthermore, wireless communication consumes the most energy in these systems. However, little attention has been paid to utilizing the feature of wireless link correlation to reduce energy consumption in these systems. This dissertation introduces system-oriented energy management research with focuses on (i) ultra-capacitor-based sustainable energy storage and sharing system design, and (ii) energy efficient network protocol (e.g., flooding protocol) design. The contributions of this dissertation are multifold. First, it is the first in-depth work to investigate the ultra-capacitor-based energy storage and sharing design. Specifically, the systems presented in this dissertation were the first to show that ultra-capacitors can be integrated into a sensor device to power the device and deliver sustainable energy to other devices inside the network. This dissertation also presents novel algorithms that improve the existed energy management techniques and provide stable, sustainable, and efficient designs. Second, we present prototype implementations and test-bed evaluations of our systems. Our results show that our systems can efficiently utilize and distribute energy inside the network. Finally, this dissertation makes a clear departure from traditional energy management which has largely proceeded in two separate directions: energy harvesting and energy conservation. Without balancing the harvested energy with the energy consumption of a device, further improvement in energy efficiency has been hindered. This dissertation proposes a design that synchronizes the device's harvested energy with its energy consumption. At the same time, the design maintains desirable properties such as distributed, low-complexity, implementable, and integrable with other energy management techniques.