A spline-based Hammerstein predistortion for 3G power amplifiers with hard nonlinearities

In this paper, we focus on the digital predistortion for the linearization of power amplifiers (PAs) with intrinsic hard nonlinearities, such as Doherty PAs, envelope tracking (ET) PAs widely explored in the third generation (3G) communication systems. Conventional Volterra-series-based predistortion with high degree nonlinearities suffers from the difficulty of real-time implement and the problem of numerical instability. Accordingly, a block-oriented Hammerstein predistortion with cubic-spline static nonlinearity is proposed, and the separable nonlinear least squares (SNLS) method is used to identify the Hammerstein coefficients, which significantly reduces the dimension of search space. A normalization convergence scheme is also presented to ensure the global optimum. The experiments are carried out on uplink single-carrier WCDMA signals, the results show that both the static nonlinear distortion and the memory effect of the PA under test are compensated significantly, and fast convergence is achieved by the normalization scheme.

[1]  Paolo Campolucci,et al.  Complex-valued neural networks with adaptive spline activation function for digital-radio-links nonlinear equalization , 1999, IEEE Trans. Signal Process..

[2]  Kasmiran Jumari,et al.  2009 International Conference on Future Computer and Communication , 2009 .

[3]  Timo Rahkonen,et al.  Measurement technique for characterizing memory effects in RF power amplifiers , 2001 .

[4]  Gene H. Golub,et al.  The differentiation of pseudo-inverses and non-linear least squares problems whose variables separate , 1972, Milestones in Matrix Computation.

[5]  Anding Zhu,et al.  Digital Predistortion for Envelope-Tracking Power Amplifiers Using Decomposed Piecewise Volterra Series , 2008, IEEE Transactions on Microwave Theory and Techniques.

[6]  Edward J. Powers,et al.  A new Volterra predistorter based on the indirect learning architecture , 1997, IEEE Trans. Signal Process..

[7]  Huihuang Chen,et al.  On the numerical stability of RF power amplifier's digital predistortion , 2009, 2009 15th Asia-Pacific Conference on Communications.

[8]  Er-Wei Bai,et al.  Convergence of the iterative Hammerstein system identification algorithm , 2004, 2004 43rd IEEE Conference on Decision and Control (CDC) (IEEE Cat. No.04CH37601).

[9]  Dennis R. Morgan,et al.  A robust digital baseband predistorter constructed using memory polynomials , 2004, IEEE Transactions on Communications.

[10]  David T. Westwick,et al.  Identification of Hammerstein models with cubic spline nonlinearities , 2004, IEEE Transactions on Biomedical Engineering.

[11]  Jangheon Kim,et al.  A Wideband Envelope Tracking Doherty Amplifier for WiMAX Systems , 2008, IEEE Microwave and Wireless Components Letters.