Sparse Array Design via Integer Linear Programming

In this paper, a design framework based on integer linear programming is proposed for optimizing sparse array structures. We resort to binary vectors to formulate the design problem for non-redundant arrays (NRA) and minimum-redundant arrays (MRA). The flexibility of the proposed framework allows for dynamic adjustment of constraints to meet various applicative requirements, e.g., to achieve desired array apertures and mitigate mutual coupling effects. The proposed framework is also extended to the design of high-order arrays associated by exploiting high-order cumulants. The effectiveness of the proposed sparse array design framework is investigated through extensive numerical analysis. A comparative analysis with closed-form solutions and integer linear programming-based array design methods confirms the superiority of the proposed design framework in terms of number of degrees of freedom (DOF) and direction of arrival (DOA) estimation accuracy.