Sparse LCMV beamformer design for suppression of ground clutter in airborne radar

Cancellation of the ground clutter received at an airborne phased array radar is an inherently two dimensional problem. Clutter returns are Doppler shifted due to platform motion forcing the use of processors that can resolve targets in both velocity (Doppler) and azimuth. Fully adaptive processors that operate in both dimensions require prohibitively large computation so that reduced adaptive dimension, or partially adaptive processors must be considered. In conventional partially adaptive linearly constrained minimum variance (LCMV) beamformer design the approach taken has been to represent the interference subspace with some reduced set of vectors, typically the eigenvectors associated with the largest eigenvalues of the interference covariance matrix. This technique does yield good performance but will not give the optimum performance for a given partially adaptive dimension. In this paper, an off-line method for selecting the “best” degrees of freedom to be retained in a partially adaptive design is presented. The sequential algorithm described selects those degrees of freedom that best minimize the beamformer output mean square error. This approach leads to a sparse structure for the transformation matrix, which when implemented in a generalized sidelobe canceller (GSC) structure results in a reduction in the computational load. This approach also allows a reduction in the required adaptive dimension as compared to the eigenvector based approach. Illustrative examples demonstrate the effectiveness of this method