Distributed Enviromental Sensor Network: Design and Experiments

Abstract

Algorithms for sensor deployment and adaptive sampling form the basis for multisensor fusion of spatio-temporal data from a wireless environmental network of deployed sensors. Derivation of sampling algorithms based on parametric methods are described. These algorithms form the basis for deployment of an array of wireless CTD (conductivity, temperature, depth) sensors to observe basic oceanographic data in Tampa Bay, Florida, USA. This distributed sensor network communicates using RF wireless 802.11b systems, and provides data in real-time to a shore observation station. In the experiments described here, five CTD sensors recorded reliable data over 25 hours. These data have been analyzed using multisensor fusion algorithms to characterize the temporal and spatial patterns. The resulting data analysis is available for integration with other observations made during these experiments, including biological and chemical variables. The approach demonstrates the ability to design and deploy a distributed sensor network that monitors real-time spatio-temporal oceanographic data, and supports further deployments that will incorporate mobile nodes capable of adaptive reconfiguration