Microsoft Word - communications-FIRST.DOC

This paper considers the application of radial basis function neural networks for antenna array systems. An overview of neural network-based direction of arrival estimation is presented, with different radial basis neural networks for signal detection stage. Following that approach, for the case where a large number of neurons can be used, a radial basis function neural network with exact solution is proposed for null-steering beamformer. The results from computer simulations will show that this network generalizes very well and null out the interference signals. The main advantage of this solution is that there is no need for network training, namely the network is only designed for a short time. Key-Wordsradial basis neural network, antenna array, null-steering beamforming.