於零填補及已知符元填補正交分頻多工系統下利用預選零子空間之通道估測

Predetermined Null Subspace-Based Channel Estimation on ZP-OFDM and KSP-OFDM

張元齊(Yuan-Chi Chang)

指導教授 ： 鐘嘉德

國立臺灣大學/電機資訊學院/電信工程學研究所/碩士(2014年)

https://doi.org/10.6342/NTU.2014.00758

摘要

由於高頻寬效應以及對抗多重路徑衰退的強韌性，正交分頻多工(orthogonal frequency division multiplexing)系統在無線寬頻通訊上被廣泛的使用，在無線及行動通訊中，由於多重路徑衰退(multipath fading channel)導致傳送信號失真，因此在通道估測在接收端的同調解調(coherent demodulation)系統中扮演非常重要的角色。
通道估測可以概括分為三類:訓練基底通道估測(training based channel estimation)，盲通道估測(blind channel estimation)，以及半盲通道估測(semi-blind based channel estimation)。實做上是最常使用通道估測是訓練基底通道估測，此系統利用與通道相等長度的訓練序列估測通道，可以利用較低的複雜度得到精準的估測，但使用之訓練序列導致頻寬效應(bandwidth efficiency)下降，相對的，盲通道估測可以在沒有任何訓練序列下估測通道，雖然保持最高頻寬效應，但估測所需的運算複雜度較高且估測性能較差。半盲通道估測則介於兩者之間，利用少許的訓練序列估測通道，以及低複雜度的運算下，達到比盲通道估測更好的估測性能。
在文獻中，盲通道估測利用的零子空間受到雜訊的影響，導致估測性能的衰退，而且在所有盲通道估測之中，存在一無解之純量不定值，只能藉由對通道的額外假設得知。
在本論文中，提出一個應用於零填補正交分頻多工系統上利用預選零子空間為基底的盲通道估測，其中的零子空間不受雜訊影響。且在已知符元填補系統上提出兩個通道估測系統，一個為利用最短之訓練序列求出純量不定值的半盲通道估測，另一個混合式訓練基底通道估測系統結合了現有的最佳訓練序列訓練基底通道估測，以及利用最佳訓練序列之預選零子空間之訓練基底通道估測系統。
根據模擬結果顯示，在零填補正交分頻多工系統上利用預選零子空間為基底的盲通道估測比文獻上可以在訊雜比不高的通道環境下提供比文獻上之盲通道估測更好的估測性能，在已知符元填補正交分頻多工上，混合式訓練基底通道估測系統利用權重調整消除了原先預選零空間通道估測系統之錯誤地板(error floor)效應，並提供比最佳訓練序列之訓練基底通道估測更精準的估測性能；最短之訓練序列求出純量不定值的半盲通道估測提供了在中等訊雜比通道下相較於文獻中半盲通道更佳的估測性能。

關鍵字

盲通道估測 ； 零填補正交分頻多工 ； 已知符元填補正交分頻多工 ； 預選零子空間 ； 均方差

並列摘要

Orthogonal frequency division multiplexing (OFDM) is a popular modulation for broadband wireless digital communication due to the high data rate and robustness against multipath fading channel. Channel estimation plays an important role in OFDM systems because each subcarrier to suffers different attenuation in multipath fading channel. Basically, channel estimation can be divided into three categories, namely, pilot-assisted channel estimation, blind channel estimation, and semi-blind channel estimation. Although pilot-assisted channel estimation is commonly used in practical communication systems, it requires additional pilot symbols to estimate channel state information (CSI) and thus results in a loss in bandwidth efficiency. In order to preserve bandwidth efficiency, blind channel estimation is recently proposed to obtain CSI without any pilot symbol, at the cost of degrading the accuracy of channel estimation estimate quality degradation. While semi-blind channel estimation emerged as a new technique which can allow significant reduction of pilot symbols while offering better performance than blind channel estimation.
In the literature, the existing blind channel estimation estimates the channel impulse response (CIR) by exploiting null subspace of received OFDM blocks which suffers from noise, and results in an unresolvable scalar ambiguity.
In this thesis, a predetermined null subspace-based channel estimation is proposed on ZP-OFDM. Two new methods are proposed on KSP-OFDM, one semi-blind based method with minimum length of known symbols which provides solvable scalar ambiguity, and the other hybrid training based combines conventional optimal training method with predetermined null subspace-based method.
Simulation results show that the proposed inverse channel based channel estimations outperform the conventional channel estimation schemes in medium signal-to-noise power ratio environment.

並列關鍵字

Blind channel estimation ； known symbol padding (KSP) ； orthogonal frequency division multiplexing (OFDM) ； predetermined null subspace ； mean square error (MSE) ； zero padding (ZP)

參考文獻

[1] H. Steendam and M. Moeneclaey, ”Different guard interval techniques for OFDM: performance comparison,” Multi-Carrier Spread Spectrum, Herrsching, Germany, May 7-9, 2007, pp. 11-24.

[2] S. Barbarossa, A. Scaglione, and G. B. Giannakis, “Performance analysis of a deterministic channel estimator for block transmission systems with null guard intervals,” IEEE Trans. Signal Process., vol. 50, no. 3, pp. 684–695, Mar. 2002.

[3] A. Scaglione, G. B. Giannakis, and S. Barbarossa, “Redundant filter bank precoders and equalizers part II: Synchronization and direct equalization,” IEEE Trans. Signal Process., vol. 47, no. 7, pp. 2007–2022, Jul. 1999.

[4] B. Su and P. P. Vaidyanathan, “Subspace-based blind channel identification for cyclic prefix systems using few received blocks,” IEEE Trans. Signal Process., vol. 55, no. 10, pp. 4979-4993, Oct. 2007.

[5] B. Su and P. P. Vaidyanathan, “Performance analysis of generalized zero-padded blind channel estimation algorithms,” IEEE Signal Process. Lett., vol. 14, no. 11, pp. 789-792, Nov. 2007.