Abstract
There is a rapidly growing need to evaluate sensor network functionality and performance in the context of the larger environment of infrastructure and applications in which the sensor network is organically embedded. This need, which is motivated by complex applications related to national security operations, leads to a paradigm fundamentally different from that of traditional data networks. In the sensor networks of interest to us, the network dynamics depend strongly on sensor activity, which in turn is triggered by events in the environment. Because the behavior of sensor networks is sensitive to these driving phenomena, the integrity of the sensed observations, measurements and resource usage by the network can widely vary. It is therefore imperative to accurately capture the environmental phenomena, and drive the simulation of the sensor network operation by accounting fully for the environment effects. In this paper, we illustrate the strong, intimate coupling between the sensor network operation and the driving phenomena in their applications with an example sensor network designed to detect and track gaseous plumes.
