Pre-Distortion for the compensation of HPA nonlinearity with neural networks: Application to satellite communications

Neural networks (NNs) are able to give solutions to complex problems in digital communications due to their nonlinear processing, parallel distributed architecture, self-organization, capacity of learning and generalization, and efficient hardware implementation. The pre-distortion being at the center of interest of this paper is one of the possible methods to compensate for HPA nonlinearities. The principle of pre-distortion is to distort the HPA input signal by an additional device called a pre-distorter whose characteristics are the inverse of those of the amplifier. In this paper, we propose a pre-distortion scheme based on a feed-forward neural network. Efficient High Power Amplifiers (HPA) present non-linearities generating amplitude and phase distortions on the HPA output signal; the proposed pre-distortion technique will reduce theses distortions. The performance of the proposed scheme is examined through computer simulations for 16-QAM OFDM signals. It is confirmed that the proposed pre-distorter with neural network consisting with one hidden layer and nine neurons gives a good performance improvement of quality of the transmission. Specifically, improvements in the reduction of the bit error rate (BER) are demonstrated for the travelling wave tube (TWT) HPA model.