A Sparse Approach to Partially Adaptive Airborne Radar

This paper is concerned with linearly constrained minimum variance (LCMV) beamforming. Partially adaptive LCMV beamformers are designed by determining a transformation which maps the fully adaptive weight space into a lower dimension partially adaptive weight space, usually so that some set of performance measures is optimised. One common method is to utilise the eigenvectors associated with the interference data covariance matrix. An iterative design technique which satisses the dual goals of mimimum output mean squared error (MSE), and reduced adaptive dimension was rst presented in 1]. This paper extends these results by considering the convergence performance of the resultant beamformer. Simulation results demonstrate that this iterative approach leads to a lower converged MSE whilst retaining simplicity in the beamforming structure.