Two-dimensional spatial-spectrum estimation of coherent signals without spatial smoothing and eigendecomposition

In array signal processing, most of the high resolution methods for direction of arrival (DOA) estimation are eigenstructure-based methods. The computational burden involved in the eigendecomposition limits the application of eigenstructure-based algorithms for real environments. A cross-covariance matrix constructed to be of full rank regardless of the coherence of the signals is suggested to avoid employing the technique of spatial smoothing in two-dimensional (2-D) spatial-spectrum estimation of coherent signals. The propagator method is used to perform the noise subspace estimation without any eigendecomposition. Furthermore, the new method can handle unknown correlated noise fields. The effects of the unknown noise are eliminated by using a rectangular planar array and a separately placed single sensor. All these properties make real-time implementation of the method possible