Adaptive Digital Predistortion of Nonlinear Systems

PhD Student 
Li Gan


 Compensating or reducing the nonlinear distortion - usually resulting from a nonlinear system - is becoming an essential requirement in many areas. In this thesis adaptive digital predistortion techniques for a wide class of nonlinear systems are presented. For estimating the coefficients of the predistorter, different learning architectures are considered: the Direct Learning Architecture (DLA) and Indirect Learning Architecture (ILA). In the DLA approach, we propose a new adaptation algorithm - the Nonlinear Filtered-x Prediction Error Method (NFxPEM) algorithm, which has much faster convergence and much better performance compared to the conventional Nonlinear Filtered-x Least Mean Squares (NFxLMS) algorithm. All of these time domain adaptive algorithms require accurate system identification of the nonlinear system. In order to relax or avoid this strict requirement, the NFxLMS with Initial Subsystem Estimates (NFxLMS-ISE) and NFxPEM-ISE algorithms are proposed. Furthermore, we propose a frequency domain predistortion technique - the Spectral Magnitude Matching (SMM) method. The ILA approach is classified into ILA-I and ILA-II approaches and the Recursive Prediction Error Method (RPEM) algorithm is proposed. In the ILA-I approach, the RPEM algorithm can well reduce the spectral regrowth and compensate the nonlinear distortion. Also, using the RPEM algorithm in the ILA-II approach can greatly improve the performance of the predistorter, compared to the traditional Least Mean Squares (LMS) algorithm. For implementation of these algorithms, General Gradient Calculation Architectures (GGCAs) are proposed for different nonlinear systems. Finally we apply these techniques for the predistortion of some nonlinear models used in practical communication systems, e.g., the parallel Wiener-type model and the memory polynomial model.  


This thesis is supervised by Gernot Kubin.