Analysis of nonlinear amplifier effects and data predistortion in QAM systems

The first part of this research is to provide a comprehensive study on the performance of quadrature amplitude modulation (QAM) in the presence of nonlinear power amplifiers. The study includes both in-band and out-of-band distortion by in corporating nonlinearity induced adjacent channel interference (ACI). Two objective functions proposed in the literature are used as performance measures in the simulations: (1) total degradation and (2) total bit energy-to-noise power density ratio for a target bit error rate (BER). Both fixed and adaptive DC bias controlled amplifiers are considered. The two objective functions result in identical designs in the case of fixed DC bias amplifiers with constant DC power but not in the case of other types of power amplifiers such as adaptive bias controlled amplifiers. In the later case the second objective function is the more useful.; The second part of this research is to study the digital baseband data predistorter (DP) with memory technique to reduce nonlinear amplifier effects in single power mode and power controlled mode QAM systems, respectively. The organization of three types of the DP with memory is introduced. The ideas to design and implement this technique are shown. The performance in terms of the above objective functions is demonstrated. In power controlled mode QAM systems, two schemes based on the DP with memory are developed: (1) look-up table with amplitude scaling and (2) interpolation/fitting-based look up table. The tradeoffs among performance, memory size and computational complexity of these schemes are investigated.