標題: 一位元高精確度之載波相位估計及高運算效率之碼相位擷取
One-bit High-Accuracy Carrier Phase Estimation and Computationally-Efficient Code Phase Acquisition
作者: 謝萬信
Hsieh, Wan-Hsin
高銘盛
Kao, Ming-Seng
電信工程研究所
關鍵字: 載波相位估計;一位元量化;反正切函數相位鑑別器;碼相位擷取;快速傅利葉轉換;carrier phase estimation;one-bit quantization;arctangent phase discriminator;code phase acquisition;fast Fourier transfrom
公開日期: 2011
摘要: 本論文首先探討高精確度之一位元載波相位估計,我們根據訊雜比(signal-to-noise ratio, SNR)選擇合適的相位鑑別器,以達到高精確度的相位檢測。在低及高訊雜比的環境中,反正切函數相位鑑別器(arctangent phase discriminator)以及雜訊補償數位相位鑑別器(noise-balanced digital phase discriminator)分別可以精確地估測載波相位,但對於中等訊雜比的應用,兩者皆有估計偏移(bias)的問題。因此,本論文提出訊雜比輔助相位鑑別器(SNR-aided phase discriminator),以解決中等訊雜比的相位檢測問題。然而,許多實際的應用無法提供訊雜比的資訊,因此我們進一步將訊雜比輔助相位鑑別器發展成同時估測相位及訊雜比的演算法。另一方面,本論文提出一個高運算效率的碼相位(code phase)擷取方法,稱為相位同步擷取法(phase coherence acquisition, PCA)。我們利用複數相量(complex phasor)擷取虛擬隨機序列(PN sequence)的碼相位,其中輸入及本地序列會先分群,接著將分群序列映射到複數相量以提升抗雜訊能力。由於相位同步擷取法主要利用複數相量之間的相位差,所以不需要複數乘法的運算,因此相位同步擷取法所需的運算量遠低於習知的快速傅立葉轉換方法。最後,本論文更進一步發展多層相位同步擷取法(multi-layer PCA),以得到更強的抗雜訊能力。
In the dissertation, we first investigate the high-accuracy one-bit carrier phase estimation. Signal-to-noise ratio (SNR) is utilized to select the proper phase discriminator to achieve high accuracy. The traditional arctangent phase discriminator (APD) and the noise-balanced digital phase discriminator (NB-DPD) can obtain accurate carrier phase for low and high SNR, respectively, but both algorithms have estimation bias in moderate SNR. Therefore, the SNR-aided phase discriminator (SNRaPD) is proposed to obtain the accurate phase. Since the SNR information may be unavailable in many applications, we further extend the algorithm of SNRaPD to jointly estimate the phase and SNR. On the other hand, we propose a computationally efficient method, termed Phase Coherence Acquisition (PCA), for PN sequence acquisition by using complex phasors. In order to combat noise, the input and local sequences are partitioned and mapped into complex phasors in PCA. The phase differences between pairs of phasors are then utilized for code phase acquisition, and thus complex multiplications are avoided. The computation load of PCA is much less than that of the conventional fast Fourier Transform (FFT) method. Finally, the multi-layer PCA is developed to enhance noise-robustness.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT079313807
http://hdl.handle.net/11536/40517
Appears in Collections:Thesis


Files in This Item:

  1. 380701.pdf