Adaptive Compensation of Nonlinearity in High Power Amplifier by Support Vector Machine

Abstract

This paper is concerned with a new adaptive predistortion method for compensation of linear and nonlinear distortions of a high power amplifier (HPA), which is frequently used in the orthogonal frequency division multiplexing (OFDM) communication scheme and mobile communication systems. The proposed nonlinear method is based on the support vector machine (SVM), which can construct an inverse system of the nonlinear amplifier through learning using a given number of training samples. An on-line adaptive algorithm for updating the parameters of SVM is also investigated to reduce its computational burden of the batch-processed optimization. By comparing with a case without predistortion and ordinary table-lookup scheme, it is shown that the proposed data-based approach can improve the suppression of the power spectrum of the amplifier output outside the input frequency band caused by the nonlinear distortions.