Shaped Power Pattern Synthesis of Linear Sparse Arrays with A Novel Hybrid Method

This paper proposes a hybrid algorithm for the synthesis of linear array with shaped beam and directivity gain constraints. The static and dynamic convex optimization(SADCP) is used to achieve a high performance on sidelobe level(SLL), null level, directivity gain and excitation control. In order to reduce the number of elements, matrix pencil method is used to get a sparse solution based on the results given by SADCP. Numerical examples are provided to show universality and high efficiency of the proposed method.

I. Introduction

Various methods for optimizing the performance of arrays have been widely used [1], such as genetic algorithm(GA), deep learning, Lagrange [2], matrix pencil method(MPM) [3] and Convex optimization is a very important method of them. Generally speaking, convex optimization is mainly divided into two kinds, the first one which does not change the solution space is called static convex optimization, and the other is called dynamic convex optimization [4]. Both static and dynamic convex optimization has been used for array synthesis in [5], and MPM has a good performance for sparse synthesis [3] [6].