論文提要及內容： 對於高資料傳輸率無線通訊系統而言，在線性調變技術中擁有好的頻譜效率是相當吸引人的。然而，此類系統的浮動波包(fluctuating envelopes) 卻結合了來自高功率RF放大器的非線性現象，以致於造成QAM調變訊號的扭曲效應(warping effect)，進而嚴重影響傳輸品質。為了有效消除傳輸端中的warping effect，本論文將使用於基頻操作的data predistorter作為所需的補償器。其中，我們使用多層類神經網路系統來當作predistorter中所使用的nonlinear filter，它將被訓練成為高功率放大器響應的反函數，並且進行基頻資料預先扭曲的非線性補償。為了實現此inverse filter，本論文中使用了多層感知器架構配合複數倒傳遞演算法 (Complex Backpropagation，CBP)類神經網路，並與最小均方演算法(Least Mean Square，LMS)與實數倒傳遞演算法(Real Backpropagation，RBP)類神經網路的效能比較。除了各種演算法的介紹外，為了論文的完整性及一致性，本論文將從基本的類神經網路來開始討論，最後再將電腦模擬的結果附上以說明各種演算法的性能比較。 Abstract of thesis: The good spectral efficiency of linear modulation techniques makes them attractive for use in high date rate digital radio system. Nevertheless, the fluctuating envelopes of such systems combined with the nonlinear nature of the high power RF amplifiers commonly. The warping effect caused by the high power amplifier will seriously degrade the transmission quality of QAM modulated signals. In order to suppress warping effect, one possibility is to use data predistorter operating at baseband as a compensator. In this case, we present a preliminary implementation of a data predistortion system using a multilayer perceptron neural network which forms an adaptive nonlinear filter whose response approximates the inverse function of the HPA nonlinearity. The neural network utilized in this work is a multilayer perceptron using Complex Backpropagation(CBP) algorithm to improve the performance of Least Mean Square(LMS ) algorithm and Real Backpropagation (RBP) algorithm.
