Intellectual Merit. Wireless sensor networks include both static and actuated devices and offer a promising solution for applications ranging from monitoring of diverse environmental phenomena to surveillance for national security. In many applications, these networks are deployed to monitor an underlying physical phenomenon over time and space, such as estimating a contaminant flux in water. Since the sensor devices operate unattended and with constrained energy reservoirs, energy efficiency becomes critical and directly determines the system lifetime. In order to achieve the simultaneous objectives of sensing fidelity and energy efficiency, it is necessary to develop strategies for selecting those sensor devices that contribute most effectively to the desired monitoring fidelity and to select other nodes for an inactive and low power standby state. The goals and intellectual merits of the proposed research are threefold: (i) to develop distributed methodologies to select a subset of sensor nodes and sensor measurement locations that can meet desired estimation accuracies; (ii) to test the developed methods experimentally; and (iii) to apply the methods to a critical application in water resource management such as accurately measuring contaminant flux in water.
Broader Impact. The research has implications on the information technology infrastructure of the nation by enhancing the ability of wireless networks to sample and extract information from geographically scattered nodes subject to energy, communication, and time constraints. The research also addresses the demand for accurate sensing coverage in water resource management systems. And it helps train students in an area of fundamental relevance to future networking systems.
