Single-Hydrophone Model-Based Passive Sonar Source Depth Classification

This paper introduces a source depth classification technique applicable to passive low-frequency narrowband sonar signals received by a single hydrophone in shallow water. The classifier is based upon Rice probability density functions with model-derived parameters for the received amplitude of a tonal signal that has been modulated by propagation through the ocean. Differences in modal excitation due to source depth result in different probability density functions of received amplitude that allow for source depth classification. The performance of the technique is demonstrated using data from the SWellEx-96 experiment.