Sector interpolation of non-uniform arrays for efficient high resolution bearing estimation

Bearing estimation is a fundamental array processing task for which attractive algorithms exist when the array is linear and uniformly-sampled. Since circumstances often require the use of an irregular two-dimensional array, the problem of interpolating a synthetic linear, uniformly sampled array from the real array is considered. Accurate interpolation is achieved by interpolating several synthetic arrays, each of which represents the real array over a limited sector of bearing angles. The utility of the method is demonstrated through a design example and simulation using a circular array and an eigenvector-based bearing estimator for linear, uniformly-sampled arrays.>