engleski

Passive Detection of Vessels

The paper treats the statistical optimum passive detection of the underwater noise (signal) of vessels in the presence of deep-sea ambient noise (noise). The underwater noise of the vessel is assumed to be a Gaussian stochastic process composed of a sum of two statistical independent parts. One part is due to cavitation phenomenon of the vessel propeller(s) and the other part is due to a finite number of degenerate tones of the vessel's operating machinery, mechanisms and propeller(s). The deep-sea ambient noise is assumed to be a Gaussian stochastic process as well. The paper considers the statistical optimum detection of such a complicated signal by means of the likelihood ratio statistical test and the Neyman-Pearson criterion. This formulation includes the statistical signal detection problem in which both the signal and the noise are Gaussian stationary stochastic processes with non-white power spectral densities. The signal has a complicated power spectral density. It consists of a wide-band component due to cavitation and of a finite number of narrow-band components due to degenerate tones. This mathematical model of the signal power spectral density is the basis of the optimum detection algorithm synthesis. The statistical optimum detection is considered in the frequency domain.

nije evidentirano

nije evidentirano