ABSTRACT
Parameters measured by sensor networks usually fall into three categories: environmental parameters such as wind speed, temperature, or the presence of some chemical agent; target features used for classification; and estimates of position and velocity along target trajectories. The quality of the local velocity determination algorithms is based in part on accurate determination of when a test vehicle is at its closest point of approach (CPA) to any given sensor node. The CPA table contains records of the sensor, the vehicle causing the peak, the node associated with the peak, the peak time, and, for acoustic and seismic sensors, the maximum energy and amplitude of the signal. The source is said to be in the near field, and is modeled as a single signal delayed by different amounts arriving at each sensor. When sources are far from the sensor network, the acoustic data can be thought of as arriving in an acoustic plane.
