Decision-Directed Detection and Spectral Probability Estimation

摘要

Over the past several years many contributions have been made to the problem of detecting underwater acoustic signals and estimating signal parameters such as time of arrival, frequency of arrival, etc. In the previous research effort, the spatial detection problem was re-examined from a decision-directed point of view, and a methodology was presented to estimate the time-varying joint prior spatial distribution of the signal. In the current research effort, attention is directed toward improving detection of temporal signals according to the frequency content of the signal. The source environment characterizes the presence of absence of the various sources (both signals and noise) as they evolve in frequency and time. It may not be assumed that this environment is stationary in either frequency content or time, hence it is necessary to characterize and track the nonstationary behavior of this environment.