High resolution two-dimensional DOA estimation using artificial neural networks

This paper presents an application of artificial neural network (ANN) based smart antenna in two-dimensional direction-of-arrival (2-D DOA) estimation. Data model based on the uniform rectangular array (URA) configuration is utilized for ANN modeling. It is shown that appropriately trained ANNs can be successfully used to localize narrowband sources in both the azimuth and elevation planes. In order to increase the angular resolution of 2-D DOA estimation a Two-Stage Sectorized Neural Network (TSS-NN) model is presented. By avoiding complex calculations the proposed approach is suitable for real-time applications demonstrating the accuracy that is comparable to the super-resolution 2-D MUSIC algorithm.