Synthesis of Sparse Antenna Arrays via Iterative Convex Optimization

The design of sparse antenna arrays is realized by a novel and effective method based on convex programming in this paper. By the use of synthesis model in terms of thinned arrays, a novel binary quadratic optimization problem is established for design of sparse array, where the required antenna element number satisfying user-defined constraints (peak sidelobes and directivity) is fully minimized. To reduce computational cost, an effective optimization approach combining the reweighted
$l_{1}$
-norm minimization with the element location expansion and contraction is proposed. By comparing to fully populated half-wavelength arrays, a numerical example is presented to assess the advantages of the described approach.