Wideband STAP (WB-STAP) for passive sonar

This paper addresses an innovative method for passive sonar signal processing where it is required to suppress a field of moving acoustic interferes while simultaneously enhancing the signal from a weak moving source. Motivated by the Space-Time Adaptive Processing (STAP) technology for similar radar applications, we propose a solution to the passive broadband sonar problem using Wideband Space-Time Adaptive Processing (WB-STAP) in the frequency-space domain. The overall objective is to develop models of the source/receiver dynamics directly into the problem formulation and analyze time varying nature of the data vector and the covariance matrix by operating in the spatio-frequency domain. This allows STAP-like formulation for the broadband passive sonar in the space-frequency domain. In particular, by employing subaperture techniques and exploiting the underlying model using eigen-subspace methods it is possible to obtain optimum adaptive weight vectors that require significantly less number of data samples compared to conventional methods. In addition, this formulation allows simultaneous tracking of all targets present in the field of view by jointly estimating their arrival angles and Doppler parameters.