Distributed space-time communication for sensor networks

Abstract

A key bottleneck in the vision of large numbers of networked sensor nodes is the energy-efficient collection of data from these nodes. In this paper, we provide two complementary approaches to this problem, both of which are fundamentally different from methods currently advocated in the literature: Virtual radar: In this method, complexity is moved to a collector node. The sensor nodes simply respond in precisely timed fashion to a beacon sent by the collector (e.g., an aircraft or vehicle), enabling the collector to use signal processing techniques similar to synthetic aperture radar (SAR) in order to construct an "image" of the activity in the sensor network. Distributed beamforming: In this technique, local clusters of nodes agree upon the data to be sent to a remote location, and coordinate their transmissions so as to form a beam in the desired direction. It is shown that even with moderate uncertainties in the locations of the sensor nodes, it is possible to get large increases in range for the same transmitted power.