Signal threshold estimation in a sensor field for undersea target tracking

This paper addresses the problem of online surveillance of undersea targets moving over a deployed sensor field. A real-time algorithm has been formulated to estimate the detection threshold based on the ensemble of sensor time series data collected from the track of a moving target. The probabilistic-state-machine-based algorithm is optimal in the sense of weighted linear least squares. The algorithm has been tested with sensor data from several tracks on a simulation test bed.