【摘要】：功率放大器是無(wú)線通信系統(tǒng)中的必要組成部分,它的非線性失真除了會(huì)造成信號(hào)帶外的頻譜擴(kuò)展,引起鄰道干擾外,還會(huì)造成帶內(nèi)信號(hào)失真,提高誤碼率。本文對(duì)高功率放大器的線性化技術(shù)進(jìn)行了系統(tǒng)的研究。首先介紹了功率放大器的非線性特性和常見的各種線性化技術(shù),其中數(shù)字基帶預(yù)失真技術(shù)是校正效果最好、最容易實(shí)現(xiàn)和最具發(fā)展?jié)摿Φ木�性化技術(shù)。在數(shù)字基帶預(yù)失真技術(shù)中,對(duì)當(dāng)前出現(xiàn)的各種預(yù)失真方案進(jìn)行了分析總結(jié),并分別提出了針對(duì)無(wú)記憶功放和有記憶功放的預(yù)失真方法。 在無(wú)記憶功放預(yù)失真技術(shù)中著重研究了查找表預(yù)失真技術(shù)與預(yù)失真自適應(yīng)算法,文中分別使用弦截法與牛頓迭代法自適應(yīng)算法結(jié)合查找表預(yù)失真技術(shù)構(gòu)建了預(yù)失真系統(tǒng)并仿真驗(yàn)證了兩種不同算法的性能。然而弦截法的收斂精度不夠高、穩(wěn)態(tài)誤差比較大,針對(duì)此點(diǎn)不足文中使用插值法代替差商法求導(dǎo)方式對(duì)弦截法進(jìn)行了改進(jìn),仿真驗(yàn)證該改進(jìn)算法可以在原弦截法基礎(chǔ)上使得信號(hào)帶外擴(kuò)展降低6dB。在目前的通信系統(tǒng)中,OFDM等寬帶信號(hào)的應(yīng)用使得放大器的記憶效應(yīng)不能忽略,所以文章對(duì)記憶功放的預(yù)失真技術(shù)也進(jìn)行了詳細(xì)的研究,首先介紹了記憶功放的行為模型,針對(duì)不同的模型介紹了兩種不同的記憶功放預(yù)失真系統(tǒng),記憶多項(xiàng)式預(yù)失真系統(tǒng)與Hammerstein預(yù)失真系統(tǒng)。并對(duì)Hammerstein預(yù)失真系統(tǒng)以及遞歸最小二乘法自適應(yīng)算法進(jìn)行了重點(diǎn)研究,遞歸最小二乘算法有著非常好的收斂特性,然而其固定的遺忘因子限制了其性能的進(jìn)一步提高,對(duì)此文章引入了可變遺忘因子的理念,在算法復(fù)雜度稍有增加的前提下可使的遺忘因子能夠隨著系統(tǒng)特性的變化而變化,這樣就使得改進(jìn)后算法不僅具有很快的收斂速度而且能夠保持很小的穩(wěn)態(tài)誤差。在信號(hào)功率譜密度的仿真中可使得帶外擴(kuò)展進(jìn)一步降低7dB之多。
[Abstract]:Power amplifier is a necessary part of wireless communication system. Its nonlinear distortion will not only cause spectrum spread out of the band and cause adjacent channel interference, but also cause in-band signal distortion and improve bit error rate (BER). The linearization technology of high power amplifier is studied systematically in this paper. Firstly, the nonlinear characteristics of power amplifier and various linearization techniques are introduced. Among them, digital baseband predistortion is the most effective, easy to realize and the most potential linearization technology. In the digital baseband predistortion technology, the current predistortion schemes are analyzed and summarized, and the predistortion methods for memoryless and memorized power amplifiers are proposed respectively. In the memoryless power amplifier predistortion technique, the look-up table predistortion technique and the predistortion adaptive algorithm are studied. In this paper, the predistortion system is constructed by using the chord truncation method and the Newton iterative adaptive algorithm combined with the look-up table predistortion technique, and the performance of the two different algorithms is verified by simulation. However, the convergence accuracy of the chord truncation method is not high enough, and the steady-state error is relatively large. The interpolation method is used instead of the differential quotient method to solve the problem in this paper. The simulation results show that the improved algorithm can reduce the out-of-band spread of the signal by 6 dB on the basis of the original chord truncation method. In the present communication system, the memory effect of amplifier can not be ignored because of the application of wideband signal such as OFDM, so the predistortion technology of memory power amplifier is studied in detail in this paper. Firstly, the behavior model of memory power amplifier is introduced. Two kinds of memory amplifier predistortion system, memory polynomial predistortion system and Hammerstein predistortion system are introduced for different models. The Hammerstein predistortion system and the recursive least square adaptive algorithm are studied in detail. The recursive least squares algorithm has a very good convergence property, but its fixed forgetting factor limits the performance of the algorithm. In this paper, the idea of variable forgetting factor is introduced, and the forgetting factor can be changed with the change of system characteristics when the complexity of the algorithm is slightly increased. In this way, the improved algorithm not only has a fast convergence rate, but also can maintain a very small steady-state error. In the simulation of signal power spectral density, the out-of-band spread can further reduce the number of 7dB.
【學(xué)位授予單位】：遼寧工程技術(shù)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2011
【分類號(hào)】：TN722.75

【參考文獻(xiàn)】

相關(guān)期刊論文 前7條

1 高鷹,謝勝利;一種變步長(zhǎng)LMS自適應(yīng)濾波算法及分析[J];電子學(xué)報(bào);2001年08期

2 南敬昌;劉元安;杜海清;;非線性功放記憶效應(yīng)分析與仿真[J];遼寧工程技術(shù)大學(xué)學(xué)報(bào);2007年02期

3 錢業(yè)青,姚天任;記憶非線性功率放大器的神經(jīng)網(wǎng)絡(luò)預(yù)失真[J];計(jì)算機(jī)工程與應(yīng)用;2004年21期

4 郭天一;廉保旺;鄒曉軍;;一種具有快速跟蹤能力的改進(jìn)RLS算法研究[J];計(jì)算機(jī)仿真;2009年08期

5 錢業(yè)青;;一種高效的用于RF功率放大器線性化的自適應(yīng)預(yù)失真結(jié)構(gòu)[J];通信學(xué)報(bào);2006年05期

6 陳星;王召巴;;超線性功率放大器的研制[J];微計(jì)算機(jī)信息;2007年32期

7 劉俊坊;南敬昌;惠曉威;;基于自適應(yīng)預(yù)失真器的BLMS算法研究[J];微計(jì)算機(jī)信息;2010年22期

相關(guān)博士學(xué)位論文 前2條

1 馮永生;預(yù)失真射頻功率放大器的研究[D];北京郵電大學(xué);2007年

2 金哲;寬帶通信中有記憶射頻功率放大器的建模與預(yù)失真方法[D];浙江大學(xué);2007年

相關(guān)碩士學(xué)位論文 前10條

1 虞強(qiáng);射頻功率放大器的數(shù)字自適應(yīng)基帶預(yù)失真技術(shù)[D];上海交通大學(xué);2007年

2 楊勇;OFDM系統(tǒng)線性功率放大器基帶自適應(yīng)預(yù)失真研究[D];電子科技大學(xué);2007年

3 王強(qiáng);射頻功率放大器預(yù)失真技術(shù)研究[D];同濟(jì)大學(xué);2007年

4 林曉晨;功率放大器的數(shù)字基帶預(yù)失真技術(shù)研究[D];大連理工大學(xué);2007年

5 吳靜;OFDM系統(tǒng)中的預(yù)失真技術(shù)研究[D];西安電子科技大學(xué);2008年

6 邢s，

本文編號(hào)：2263283