Direction-of-arrival estimation using eigenstructure methods for maneuvering arrays

A general solution for extending subspace techniques to maneuvering platforms is proposed. The method involves a matrix transformation that uses an approximate guess of the DOAs (directions of arrival) and knowledge of the platform orientation corresponding to each data set to transform the available data to a consistent datum subspace. Any subspace estimation technique, such as the MUSIC, ESPRIT, or maximum likelihood methods, can then be used to arrive at the DOA estimation solution. If necessary, the solution obtained can be iteratively improved by using the estimated DOAs at each iteration for deriving the transformation needed in the next iteration. Apart from maneuvering arrays, the proposed method is of interest in compensating for unavoidable platform motion arising due to large amounts of pitch, roll, or yaw