標題: OFDM系統之同步與通道估測技術研究
Synchronization and Channel Estimation Techniques for OFDM Systems
作者: 謝孟翰
Hsieh, Meng-Han
魏哲和
Che-Ho Wei
電子研究所
關鍵字: 同步;通道估測;正交分頻多工;Synchronization;Channel Estimation;OFDM
公開日期: 1997
摘要: 正交分頻多工 (orthogonal frequency-division multiplexing, OFDM) 技術可有效地解決通訊傳輸系統中的一些問題,如頻際干擾、多重 路徑衰落等。然而,在 OFDM系統中一個主要的瓶頸在於信號的同步方面 ,尤其是對於載波頻率同步的要求很高。因此,本論文主要的目的就在於 探討 OFDM 系統中之同步與通道估測技術,並提出一些低複雜度的架構。 本論文分為以下三大主題: 第一部份我們提出一個快速、低複雜度的 同步架構。此同步架構主要是利用了OFDM系統中隔離區間 (guard interval) 的特性,同時解決了框位同步與載波頻率同步的問題。在此一 同步架構中,由於我們只使用接收信號的正負號來估測框位以及頻率漂移 量,因而大量地降低了運算以及硬體的複雜度。在效能分析方面,信號同 步的準確度則可藉由平均估測值或是追蹤迴路的方式來提高。 第二部 份我們探討 OFDM 系統中的頻率擷取問題。在一般 OFDM 系統的頻率估測 演算法中,可修正的偏移量通常只能達到次通道頻寬的 +-1/2。因此,在 實際的應用上我們仍需要一個頻率擷取架構來修正較大的頻率偏移量。在 此我們利用了導頻信號以及OFDM系統中次通道頻率響應的特性來擷取頻率 偏移量。此一頻率擷取架構可與第一部份所提出的同步架構搭配使用,將 可修正的頻率偏移量提高至近乎信號頻寬的一半。 本論文的第三部份 研究 OFDM 系統的通道估測架構,並設計一高效能、低複雜度的通道估測 架構。此一通道估測架構採用梳型的導頻信號,適用於有頻率選擇性、衰 落現象的通道。我們所提出的通道估測架構可以分為兩部份:導頻信號估 測以及通道內插法。在導頻信號估測方面,我們採用了最小均方誤差的演 算法,有效地抑制了頻際干擾以及高斯雜訊。同時,藉由低階近似法的技 術,我們可以大幅降低最小均方誤差法的運算複雜度。在通道內插法方面 ,我們研究了線性及二階多項式內插法。我們也研究了定時誤差對通道內 插法的影響,同時提出了以頻域相位補償的方式來解決定時誤差的方法。 由我們的模擬得知,結合了最小均方誤差估測導頻信號法以及相位補償線 性內插法的通道估測架構,可以在多重路徑衰落的環境下提供可靠的通道 資訊。 Orthogonal frequency-division multiplexing (OFDM) is an effective transmission scheme to cope with many impairments, such as cochannel interference, serve multipath fading and impulsive parasitic noise. However, a high accuracy in synchronization is required in OFDM systems. In this thesis, we present low-complexity synchronization and channel estimation algorithms for OFDM systems. The thesis is divided into three parts. The first part presents a fast low-complexity synchronization algorithm for OFDM systems. This algorithm is based on utilizing the guard interval in OFDM signals, and the frame synchronization and the frequency offset estimation are considered simultaneously. The implementation can be simplified by only using the sign bits of the in-phase and the quadrature components of the received OFDM signal for frame synchronization and frequency offset compensation. By using a closed tracking loop or averaging the estimate over several symbols, the accuracy is significantly increased. In the second part, a frequency acquisition algorithm for OFDM systems is presented. Many frequency offset tracking algorithms for OFDM systems suffer from the subcarrier ambiguity, and the applicable range is limited to +-1/2 of the subcarrier spacing. To solve the subcarrier ambiguity problem, we present an frequency acquisition scheme based on pilot signals and the special characteristic of subchannel responses in OFDM systems. The proposed acquisition algorithm can be used in cooperate with several frequency detection algorithms for OFDM systems, including the frequency offset estimation algorithm presented in the first part. In the third part, a low-complexity channel estimation method for OFDM systems is presented. The channel estimation algorithms is based on the comb-type pilot subcarrier arrangement, which is suitable for OFDM systems over fading channels. The proposed algorithm can be divided into two stages: the pilot signal estimation and the channel interpolation. For pilot signal estimation, estimators based on the minimum mean squared error (MMSE) criteria are studied. Owing the the MMSE estimate of pilot signals, the inter-carrier interference and additive white Gaussian noise are reduced considerably. The computational complexity of the MMSE estimator is reduced by applying the low-rank approximation technique. For the channel interpolation algorithm, linear and second-order polynomial interpolation are studied. The impact of timing errors on the channel interpolation is also investigated. In addition, phase compensators before and after the channel interpolation are presented to combat the timing errors. According to our simulations, the channel estimator based on MMSE estimate of pilot signals and phase compensated linear interpolation provides a reliable channel estimate for OFDM systems over fading channels.
URI: http://140.113.39.130/cdrfb3/record/nctu/#NT860428005
http://hdl.handle.net/11536/62984
Appears in Collections:Thesis