CHARACTERIZATION OF STOCHASTIC PROPAGATION AND SCATTERING VIA GABOR AND WAVELET TRANSFORMS
An application to remote acoustic sensing that remains unexploited is measuring acoustic scattering and spreading effects with wideband, coherent signal processing techniques. Such techniques allow distributed objects, such as a layer of scatterers due to bubbles or biological particles, and first order time variations in an ocean channel to be estimated. This paper presents narrowband and wideband methods for characterizing stochastic propagation and acoustic scattering in a time-varying ocean in terms of spreading functions. It is shown that the Gabor transform is the natural transform for estimating the narrowband spreading function, and the wavelet transform is the natural transform for estimating the wideband spreading function. Both techniques of characterization use a correlator processing structure in a monostatic transmitter/receiver configuration to estimate the spreading function. The narrowband and wideband spreading functions characterize the distribution of scatterers in range and velocity (time and frequency) in a propagation channel. It is shown that the wideband formulation follows directly from a physical derivation. Moreover, wideband processing removes many of the narrowband restrictions and allows first order time variations, caused by inhomogeneities and relative motion in the ocean channel, to be processed. In addition, wideband techniques allow for increased time intervals and, therefore, increased energy transmission when the transmitter is peak-power-limited. Thus, weak scatterers that may have been unidentified with narrowband techniques may be identified with the wideband methods. Numerical examples for wideband characterization of a distributed scatterer are presented.