Estimation of the directions of arrival of spatially dispersed signals in array processing

The problem of estimating the DOA of spatially distributed signals is examined. A mathematical model is first established by making some reasonable assumptions. The correlation matrix of a distributed signal is then derived. The important properties of the correlation matrix are studied, revealing that even if the matrix is of full rank (being equal to the number of sensors), which renders conventional high resolution array processing methods inapplicable, the dimensionality of the signal subspace can be approximated to a number usually much smaller than the number of sensors. From the observation, the quasi-signal and noise subspaces are identified, and, utilising the orthogonality of the signal and noise subspaces, an algorithm (DISPARE) has been developed. Analytic studies and numerical evaluations are carried out to examine the performance of the algorithm under various environments and show that it is indeed effective.