Optimal Design of Nonredundant Sparse Arrays With Noninteger Distance Between Elements

Direction-of-arrival (DOA) estimation using a sparse array has recently attracted considerable interest as an approach for increasing the number of detectable sources and reducing the adverse effects of mutual coupling between array elements. Nonredundant arrays (NRAs) are among the most common sparse array configurations. Each element in a conventional NRA is usually located at an integer multiple of the half-wavelength. This article proposes a method for designing optimal NRAs (ONRAs), where array elements can be placed at locations which are integer multiples of a fraction of the half-wavelength. As there are a limited number of available ONRAs for a desired number of array elements, the proposed method increases the number of available ONRAs. In addition, as the proposed design method has more flexibility in selecting the places of array elements, this enables us to obtain arrays that have a larger minimum distance between array elements, and therefore, the NRAs obtained by the proposed method can improve DOA estimation accuracy in the presence of mutual coupling. In addition to the ONRAs with minimum array aperture, this article also examines ONRAs with the desired aperture (ONRAda), ONRAs with reduced mutual coupling (ONRAmc), and hybrid ONRAs (ONRAhybrid). Simulation results show that, in the presence of mutual coupling, the NRAs obtained by the proposed method can outperform existing NRAs in estimating source DOAs.