Array pattern synthesis using neural networks with mutual coupling effect

Abstract

In the paper a neural network algorithm is presented for the synthesis of array patterns in the presence of mutual coupling. The algorithm is based on wavelet activation functions and is tested on a linear dipole array with a shaped pattern. The excitations and the lengths of the dipoles in the array are determined. The architecture of the neural network is discussed and simulation results are presented. The proposed algorithm makes the synthesis procedure more practical and accurate compared to the synthesis technique without mutual coupling considerations.