Orthogonal Neural Network for Nonlinearities Reduction of Multicarrier Transmitters

Abstract

An orthogonal neural network with Hermite activation function to reduce nonlinear distortion introduced by a traveling wave tube amplifier (TWTA) over a Multicarrier Transmitter is proposed. The neural network has the fixed input weights and only linear output weights are estimated using the Moore-Penrose generalized inverse method. Computer simulation results show that proposed neural network achieves excellent performance in nonlinearities reduction caused by TWTA. In addition, proposed scheme has a low complexity and fast convergence speed.