Jacob George
Ronald A. Wagstaff
Naval Res. Lab., Code 7176, Stennis Space Center, MS 39529
Recently the topic of wavelet analysis has received considerable attention from physicists, engineers, and mathematicians [C. K. Chui, An Introduction to Wavelets (Academic, New York, 1992)]. One well-known feature of a wavelet transform is the choice of a flexible time window which automatically narrows when observing high-frequency phenomena and widens when studying low-frequency environments, in contrast to a fixed window in traditional Fourier transforms. Alternately, the wavelet transform can be viewed as a two-parameter representation of a signal. These features have been used with advantage in time-frequency analyses [Badiey et al., ``Shallow water acoustic/ geoacoustic experiments at the New Jersey Atlantic Generating Station site,'' J. Acoust. Soc. Am. 96, 3593--3604 (1994); Drumheller et al., ``Identification and synthesis of acoustic scattering components via the wavelet transform,'' J. Acoust. Soc. Am. 97, 3649--3656 (1995)]. Fluctuation based processing which has recently been developed at NRL, takes advantage of the time fluctuations of the signal to improve the signal/noise ratio and to enhance signal detection for several frequencies. The improvements in such processing due to the use of wavelet representation will be discussed. [Work supported by ONR.]