Subspace Detection and Enumeration of Underwater Acoustic Sources
G. V. Anand, P. V. Nagesha, Sanjeev Gurugopinath · 2018
We consider the problem of narrowband detection and enumeration of underwater acoustic sources by a hydrophone array. Since the array signal vector is not known due to the unknown location of the sources, detection is based on a generalized likelihood ratio test (GLRT). A subspace detection technique viz. the modal subspace detector (MSD) was developed recently for solving this problem. This detector is based on the fact that the array signal vector in a range independent ocean lies in the modal subspace of the underwater acoustic channel. In this paper, we propose two novel subspace detectors, viz. the embedded subspace detector (ESD) and the signal subspace detector (SSD). These new detectors exploit the fact that the array signal vector also lies in the signal subspace of the covariance matrix of the array data vector. The ESD can be used only if the normal mode functions of the channel are known, while the SSD does not require this knowledge. We have also developed a recursive procedure, based on MSD, ESD or SSD, for source enumeration, i.e. estimating the number of sources. Simulation results are presented to illustrate the signal detection and source enumeration performance of all the methods.