An array processing algorithm for reconstruction of distributed targets in incompletely defined correlated noise

An algorithm for the extraction of Gaussian noise of unknown spatial correlation from data received from distributed targets at an arbitrarily configured array is developed. Knowledge of the sampling procedure is used to build a basis set representation of a spectrum constrained only by the condition that it be bandlimited. By adopting maximum a posterior (MAP) criteria it is shown that it is possible to optimally remove noise of unknown correlation from this spectrum. An application of this technique is the problem of identifying point targets in sea surface clutter environments. In this case, it is common that some knowledge is available about the noise correlation, but that this knowledge is often incomplete. This prior knowledge is used to develop a new class of objective functions that can be optimized to significantly enhance the resolution attainable through application of the conventional MAP approach.<>