Fast DOA Estimation Using Coarray Beamforming with Model Order Estimation

The paper examines the problem of direction finding (DF) using difference coarray beamforming. The employed sparse array is configured through switches that connect the front-end receivers to the respective antennas selected from those available at grid points spaced by half wavelength. In the case considered, the fully filled co-array is the well-known minimum redundancy array. The emphasis is on the application of DF techniques that most support the cognitive Perception-Action cycle by providing fast direction-of-arrival (DOA) estimation of a large number of sources relative to the array aperture. The paper modifies the fast iterative interpolated beamforming (FIIB) technique to account for the a priori known noise power as well as the Hermitian nature of autocorrelation function. We demonstrate, using Monte Carlo simulations, the modified FIIB capability of providing accurate estimates of the number of sources in the field of view.