Font Size: a A A

Research On RF Power Amplifier Digital Predistortion

Posted on:2011-11-23Degree:MasterType:Thesis
Country:ChinaCandidate:Y ZhangFull Text:PDF
GTID:2178330332969597Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
In order to improve the output power and efficiency of modern wireless communication system, RF power amplifiers is designed to work in colse to the saturation point which bring a very serious nonlinear result. Therefore, RF power amplifiers become one of major nonlinear source in the system. The nonlinear distortion of RF power amplifier creates in-band and out-band distortion.The former makes the bit error rate increase, the latter produces adjacent channel interference because of the spectrum growth beyond the signal bandwidth.which reduce the bandwidth efficiency.In order to avoid such problems, linearization technology is used to achieve the system's high linearity and high power efficiency. Among all linearization techniques, digital predistortion technology has a broad application prospect because of the high efficiency, low cost, as well as the simplist to realize.How to establish the RF power amplifier model and how to use adaptive algorithm to identify the parameters of the predistortion are studied in this paper.In the MATLAB environment, we construct three power amplifier models including the memoryless model,the Wiener modeland the Hammerstein model. Furthermore, NG iterative algorithm and LS/ SVD (Least square / Singular Value Decomposition) algorithm are used to respectively design adaptive Hammerstein predistorter. Meanwhile,LR and the LR combined with LMS algorithm are also used to design Wiener adaptive predistortorer respectively. While, NG iterative algorithm has a sensitive problem to the initial values so that LR / NG algorithm is proposed which can avoid local convergence problems effectively. With the comparison between the performance of memoryless predistorter and memory, memory predistorter power amplifier linearity can be improved up to 36dB, which has demonstrated a good performance on suppression of spectral regrowth. Also, the simulation results show that the Hammerstein adaptive predistorter using of LR / NG algorithm can improve the power amplifier linearity better which can be improved up to 10dB, in that confirms the validity and justifiability.
Keywords/Search Tags:Power amplifier, Linear technologies, Digital predistortion, Wiener model, Hammerstein model
PDF Full Text Request
Related items