Performance of Neural Network Trained with Genetic Algorithm for Direction of Arrival Estimation
Hamed Movahedi Pour, Zahra Atlasbaf, Mohammad Hakkak · 2006
Direction of Arrival (DOA) estimation has turned out to be extremely vital by reason of recent developments in Spatial Division Multiple Access (SDMA) systems. Superresolution algorithms such as the Multiple Signal Classification (MUSIC) and neural networks have been approached to carry out DOA estimation. In this paper, a Multi-Layer Perceptron (MLP) network using Genetic Algorithm (GA) method for training is proposed. The performance of the proposed network is compared with Radial Basis Function Neural Network (RBFNN) which has been considered as an effective solution to the DOA problem. It is demonstrated that by exploiting the genetic algorithm based MLP, error attributes of the estimation improve, despite the reduction of neural network size.