Fast Sidelobe Computation for Arbitrary Two-Dimensional Array for Array Sparse

Array design plays a critical role in diverse applications, including communication, radar, navigation. Sidelobes, significant influencers in array performance, are typically constrained during optimization. However, inefficient sidelobe calculations contribute to challenges, particularly in extended optimization times, especially with the increasing array sizes nowadays. To address this challenge, this paper presents a solution integrating projection preprocessing, kernel methods and neural network algorithm, ensuring fast sidelobe computation in arbitrary two-dimensional arrays. Additionally, we implement the proposed method into array sparse design. Simulation experiments demonstrate that the proposed method achieves sidelobe calculation accuracy comparable to traditional search methods. Significantly, it enhances calculation efficiency by over 38 times in both sidelobe calculation and array sparse design experiments. These results affirm the effectiveness and reliability of the proposed method for rapid sidelobe computation.