An Antenna Array Synthesis Method for Varied Aperture Lengths Based on Recurrent Neural Network

A deep learning-based antenna array synthesis method for varied aperture lengths is proposed in this article. Inspired by natural language processing (NLP), a tri-branched recurrent neural network (RNN) based on long short-term memory (LSTM) layers and an “encoder–decoder” structure is designed as the backbone architecture of the proposed model. To further enhance deep models’ comprehension of physical scenarios, this article also proposes an improved word embedding method that allows continuous real-valued parameters to be mapped into high-dimensional vectors in a manner similar to individual vocabulary terms while preserving their original “semantic” relationships. Based on the approaches mentioned above, the proposed model is built and operates in an interactive way, similar to a chatbot. Once receiving the main lobe direction and array element positions as inputs, it automatically generates amplitude-phase synthesis results that meet specified requirements, while ensuring that sidelobe levels (SLLs) are minimized to the greatest extent possible. The numerical experiments show that this method can generate patterns whose main lobe covers the range from 45° to 135° on nonuniform linear arrays (NULAs) with 15–30 isotropic elements, while controlling the SLL below 15 dB, effectively improving the directivity of the antenna array. Compared to genetic algorithms (GAs), particle swarm optimization (PSO), and deep neural network (DNN), this method demonstrates significant advantages in both synthesis performance and computational efficiency, significantly reducing the training time while achieving better radiation pattern characteristics.