Optimal design of a large dual-polarization microstrip reflectarray with China-coverage patterns for satellite communications

Abstract

A large dual-polarization microstrip reflectarray with China-coverage patterns in two operating bands is designed. To sufficiently compensate for the spatial phase delay differences in two operating bands separately, a three-layer rectangular patch element is addressed, which is suitable for the large dual-polarization reflectarray. Due to the complexly shaped areas and high gain requirements, there are more than 25 000 elements in the reflectarray, making it difficult to design, due to more than 150 000 optimization variables. First, the discrete fast Fourier transform (DFFT) and the inverse DFFT are used to establish a one-to-one relationship between the aperture distribution and the far field, which lays a foundation for optimizing the shaped-beam reflectarray. The intersection approach, based on the alternating projection, is used to obtain the desired reflection phases of all the elements at some sample frequencies, and a new method for producing a suitable initial solution is proposed to avoid undesired local minima. To validate the design method, a dual-polarization shaped-beam reflectarray with 7569 elements is fabricated and measured. The measurement results are in reasonable agreement with the simulation ones. Then, for the large broadband reflectarray with the minimum differential spatial phase delays in the operating band, an approach for determining the optimal position of the feed is discussed. To simultaneously find optimal dimensions of each element in two orthogonal directions, we establish a new optimization model, which is solved by the regular polyhedron method. Finally, a dual-band dual-polarization microstrip reflectarray with 25 305 elements is designed to cover the continent of China. Simulation results show that patterns of the reflectarray meet the China-coverage requirements in two operating bands, and that the proposed optimization method for designing large reflectarrays with complexly shaped patterns is reliable and efficient.