Abstract:
A family of processors developed at NRL [R. A. Wagstaff, ``The WISPR Filter: A method for exploiting fluctuations to achieve improved sonar signal processor performance,'' (submitted to J. Acoust. Soc. Am.)] have been shown in data analyses to yield gains in signal-to-noise ratios, and detect weak signals buried in noise. These processors take advantage of the notion that signals from submerged sources (especially those which undergo relatively fewer interactions with the surface) are relatively steady though they may be weak, but ambient noise has greater fluctuations. Simulations using synthetic data have been done to verify the validity of the above premises, and to assess the performance of the family of filters collectively called ``WISPR'' under several environments. The simulations are also compared with analyses of experimental data. The results of the simulations and the experimental analyses will be discussed. [Work supported by ONR.]