Evaluation of an MLP-based Direction of Arrival System Using Genetic Algorithm for Training

This paper considers the problem of direction of arrival (DOA) estimation of mobile users using linear antenna arrays. Radial basis function neural network (RBFNN) is a well known approach to decrease the computational complication of superresolution algorithms which are classic solutions for DOA estimation. This paper discusses the application of a multi-layer perceptron (MLP) network using genetic algorithm (GA) for training and compares the performance of this approach with RBFNN. Error attributes of direction estimations are demonstrated and compared for different angular separations of received signals.