Preliminary Analysis of Threshold Signal Detection in Ambient and Signal-Dependent Noise Environments.

摘要

Weak signal detection of signals scattered by ocean wave surfaces and reflecting bodies is examined. Optimum threshold algorithms, are determined, which contain in addition to the usual ambient background noise, signal-dependent noise generated in the scattering process. Threshold performance probabilities are formulated, in terms of the now-generalized performance parameters associated with coherent and incoherent detection. As before, a canonical theory is presented, which is invariant of specific noise statistics and signal waveforms. Both spatial and temporal processing are included, the former implicitly in may cases by means of projected beam patterns. The physico-geometrical factors of platform motion, beam patterns, distributed scattering elements and source and receiver geometries for both the monostatic and bistatic regimes are summarized. Various distributed target models are presented, including quasi-phenomenological ones capable of analytic evaluation, involving both continuous and discrete scatterers. Under the present condition of narrow-band signals and far-field geometries, the input signal waveforms are explicitly separable from these physico-geometric factors. A weak target is assumed, in weak reverberation, vis-a-vis strong ambient noise. Sufficiently detailed analytic structure is provided to achieve explicit, quantitative results, and to guide the treatment of special problems. A short list of possible next steps is included. (JHD)