Radial basis function neural network algorithm for adaptive beamforming in cellular communication systems

Abstract

A smart antenna based on a neural network implementation of the optimum Wiener solution for the problem of adaptive interference nulling using circular arrays is presented. Modem cellular satellite mobile communications systems and GPS systems suffer from different sources of interference which limit system capacity. This paper develops a fast tracking system to constantly track the mobile users, and then adapt the radiation pattern of the antenna to direct multiple narrow beams to desired users and nulls to sources of interference. The computation of the optimum weights is viewed as a mapping problem which can be modeled using a three-layer radial basis function neural network (RBFNN) trained with input/output pairs. The results obtained from this network are in excellent agreement with the Wiener solution. It is found that networks implementing these functions are successful in tracking mobile users as they move across the antenna's field of view.