Abstract
The Internet-of-Things (IoT) has the potential to revolutionize our understanding and real-time awareness about the world around us, to make our lives simpler, safer, and healthier. However, the growth of the IoT depends greatly on how efficient and cost effective the technology is. Many battery-based systems suffer from their restricted power supply lifetime, which cause regular maintenance, replacement, and downtime costs. Energy-harvesting systems can suffer from poor dependability when environmental sources supply insufficient power for their operation, which leads to poor application quality. In general, when the system’s power consumption is lowered, both the lifetime and the dependability are increased. Therefore, to support the ever growing scale of IoT device deployment, dominating factors of system power consumption must be addressed.
The goal of this work is to highlight some capabilities of modeling with respect to self-powered systems, and to illustrate the need for modeling to be further developed to aid in system design. The scale and impact of IoT devices are largely limited by the lifetime or longevity of the systems in their respective environments. From a systems perspective, modeling can aid in the realization of the design space that restricts a designer's boundaries by addressing individual components' contributions to overall system power. Further, test modes and models can be paired with statistical representations of environments to determine the functionality of a system when it is being supported solely by the environment. This is useful to help maximize the reliability and availability of deployed sensor nodes. From a sensing perspective, this work calls attention to the impact of physiology in on-body sensing. A three-part model is presented to encapsulate the impact of human-electronic interaction on ULP sensing and its accuracy. Modeling is an extremely powerful tool to aid designers in the exploration of the boundaries of current technology. When customized to specific applications, models can lend powerful insight to the design knobs that can and need to be explored for optimal solutions.
