基於拉格朗日法之適應性數位預失真方法; Lagrange Multiplier Based Method for Adpative Digital Pre-distortion

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Communication Engineering > Electronic Thesis & Dissertation > Item 987654321/48339

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/48339

Title:	基於拉格朗日法之適應性數位預失真方法;Lagrange Multiplier Based Method for Adpative Digital Pre-distortion
Authors:	儲志平;Chih-Ping Chu
Contributors:	通訊工程研究所
Keywords:	功率放大器;適應性預失真;線性化;頻帶增生;均方誤差;割線法;次梯度法;拉格朗日乘子法;誤差向量幅度;power amplifier (PA);predistortion;Lagrange multiplier method;Sub-gradient method;Secant method;mean squared error (MSE);adaptive predistorter;EVM;spectral regrowth;linearization
Date:	2011-07-27
Issue Date:	2012-01-05 14:51:55 (UTC+8)
Abstract:	因為功率放大器產生的振幅(AM/AM) 和相位(AM/PM) 非線性效應，會導致調變訊號的誤差向量幅度(error vector magnitude, EVM)擴大和頻帶外的頻譜增生，以致於干擾鄰近通道的傳輸訊號。為了有效解決此一問題，本論文研究了預失真技術。傳統上，在過去的文獻中，預失真器通常設計為一多項式模型，基於極座標法(polar form)，產生兩路的輸出訊號，分別對功率放大器的振福和相位進行線性化的補償。本篇論文先將真實功率放大器的模型表示成一多項式通式，藉由最小平方近似法(Least-square approximation method, LSA) 統一功率放大器的模型表示後，設計者可以更容易觀察諧波失真的嚴重度。再應用拉格朗日乘子法(Lagrange multiplier method)同時對功率放大器的非線性振幅和相位進行限制，期望使用單一路徑輸出的預失真訊號，就能達成高度的線性化補償。模擬結果顯示在固定的調變訊號(如：64QAM)和回退率環境下，功率放大器可以操作在接近飽和點的位置，仍舊保有高度的功率因素(power factor)。同時，帶外頻譜也可以有效的被抑制，使得接收端的錯誤率十分逼近線性的通道表現。　The nonlinear effect of a power amplifier (PA) due to the AM/AM and AM/PM characteristics usually leads to a significant error vector magnitude (EVM) and out-of-band spectral regrowth, which interferes the signals in adjacent transmission channels. To solve this problem, we study a pre-distortion scheme for the PA based on the Lagrange multiplier method. According to previous research in the literature, the PA’s pre-distorter used to utilize a polynomial model and the polar form to produce the compensating signals for AM/AM and AM/PM effects. Without following this routine work, we first establish the polynomial models for PA characteristics by the least-square approximation method. Through these models, we can easily calculate the harmonic distortion due to the PA. Then, the adaptive Lagrange multiplier method is developed to find the compensating signals for amplitude and phase. Simulation results show that the compensated PA output can achieve the high power factor even when the PA works almost close to the saturation region with a constant modulation type such as 64-ary QAM and the same back-off condition. Besides, the out-of-band spectrum can be suppressed effectively so that the bit error rate performance of the proposed pre-distortion method is almost as good as the result obtained in the linear channel.
Appears in Collections:	[Graduate Institute of Communication Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	832	View/Open

社群 sharing

Loading...