ABSTRACT
Wireless sensor networks are an emerging technology for monitoring the physical world with a densely distributed network of wireless nodes [1,2]. Each node has limited communication and computation ability and can sense the environment in a variety of modalities, such as acoustic, seismic, and infrared. In principle, sensor networks can be deployed anywhere: on the ground, in the air, or in the water. Once deployed, the nodes have the ability to communicate with each other and configure themselves into a well-connected network. A wide variety of applications are being envisioned for sensor networks, including disaster relief, border monitoring, condition-based machine monitoring, and surveillance in battlefield scenarios. Detection and classification of objects moving through the sensor field is an important task in many applications. Exchange of sensor information between different nodes in the vicinity of the object is necessary for reliable execution of such tasks for a variety of reasons, including limited (local) information gathered by each node, variability in operating conditions, and node failure. Consequently, development of theory and methods for collaborative signal processing (CSP) of the data collected by different nodes is an important research area for realizing the promise of sensor networks.
