The project focuses on the development and performance evaluation of resource allocation algorithms for wireless networks composed of nodes with rechargeable energy sources. Recent advances in energy harvesting and ultra-low-power wireless communications will soon enable the realization of such networks that will be used in disaster recovery, supply chain management, industrial control, and health care applications. Energy harvesting shifts the essence of resource allocation from prolonging the finite lifespan of a device to enabling perpetual life, thereby posing fundamentally new problems. To solve these problems, an energy management module that determines the energy consumption level is being developed. It is based on solutions to new theoretical utility maximization problems under various settings resulting from different hardware characteristics (e.g., battery and capacitor) and different harvesting environments. It will provide important inputs to protocols responsible for sleep-wake, power control, routing, and scheduling. To enable the development of the module, the goals of the project are: (i) experimentally characterize energy availability in various environments; (ii) design resource allocation algorithms for predictable, partially-predictable, and random energy sources; and (iii) study the tradeoffs between the communication and computation complexities of the algorithms, and experimentally evaluate their performance. The project will provide fundamental understanding of the design tradeoffs in networks of rechargeable devices and will bridge the gap between distributed algorithms and wireless network optimization, and the emerging field of energy harvesting. It will facilitate the operation of self sustainable networks that will have tremendous societal impacts in areas such as environmental monitoring and disaster recovery.
