Synthesis of Sparse Planar Arrays Using Off-grid CS

A novel method for the synthesis of sparse planar arrays with prescribed pattern features is presented. Unlike conventional compressed sensing (CS) approach, the active elements are not assumed to lie on the pre-specified grids, but can be randomly distributed in the given array aperture. Towards this end, the synthesis problem is formulated as a ridge regression model with the element excitations and locations all unknows. The alternative optimization strategy can be adopted herein to solve the synthesis model by estimating the layout and the excitation of the sparse planar array, simultaneously. Numerical results show the effectiveness of the proposed method in the reduction of the number of elements as well as reconstructing the desired focused pattern and flat-top pattern accurately.