No abstract

This paper develops a procedure for the optimum detection of a deterministic, finite length signal obscured by Gaussian-Markovian noise that approximates sea noise by employing the concept of a Likelihood Ratio. For the particular noise model, the optimum procedure is to use a cross-correlation detector. The variables fed to the correlation device are additive combinations of each channel's signal and its derivative. This procedure is equivalent to filtering each channel to render the background noise white and then cross-correlating the filter outputs.

A Volterra functional expansion for the logarithm of the likelihood ratio for detection of a memoryless nonlinear functional of a random process in Gaussian white noise

No abstract

No abstract