Signal Processing and Performance Analysis for the Layered Alamouti Space-Time Block Coding MIMO System

Signal Processing and Performance Analysis for the Layered Alamouti Space-Time Block Coding MIMO System

邱俊甯(Chun-Ning Chiu)

指導教授 ： 劉宗憲

國立中正大學/工學院/通訊工程學系/博士(2013年)

摘要

Layered Alamouti space-time block coding (STBC)是一個結合了spatial multiplexing與STBC的MIMO系統，本篇論文將針對此系統下的訊號處理以及錯誤率效能做深入的分析與討論。由於Alamouti STBC正交編碼的特性，使得系統的等效通道模型亦具有正交性，利用此特性我們得以發展出低複雜度的區塊化矩陣運算。相較於傳統的Householder QR分解，利用我們所提出的區塊化 Householder transformation對系統的等效通道做QR分解時大約可以節省75%的運算量。我們亦提出區塊化複數Givens rotation演算法，相較於傳統的Givens rotation，我們所提出的區塊化旋轉方式大約可以節省57%的旋轉次數。此外，我們也將區塊化的QR分解延伸至區塊化sorted QR分解，使得接收端的檢測機可以在預處的階段得到經重新排列後的等效通道矩陣QR分解。
為了更近一步改善此系統的錯誤率效能，我們將Tomlinson - Harashima precoding (THP)技術應用於此系統中，因而提出三種適用於layered Alamouti STBC系統前置編碼的方法，分別為區塊化MMSE THP、區塊化least mean square error (LMSE) MMSE THP與實數LMSE MMSE THP。從模擬的結果中顯示，若在layered Alamouti STBC系統中使用我們所提出的前置編碼方法則可以大幅地降低系統錯誤率。
最後，基於matrix variate distribution所得到的結果，且在獨立且同分佈的Rayleigh fading通道環境中，我們推導出此系統在zero-forcing successive interference cancellation (ZF-SIC)檢測方法下的系統錯誤率，由模擬的結果顯示，我們所推導出來的理論錯誤率曲線與Monte Carlo模擬出錯誤率曲線吻合。

關鍵字

Tom linson -Harashima precoding ； 區塊化複數 Givens rotation ； layered Alamouti STBC ； 區塊化HouseholderQR分解 ； space-time block coding (STBC) ； 訊號錯誤率

並列摘要

This dissertation investigates the signal processing and performance analysis for the layered Alamouti space-time block coding (STBC) system, which
is the combination of the spatial multiplexing and STBC systems. Due to the orthogonality of the Alamouti STBC, the equivalent channel matrix of the
signal model also has the orthogonal property that is further utilized to develop the block matrix computation schemes. We propose the block-wise
Householder based QR decomposition (BH-QRD) for the equivalent channel matrix. Compared with the computational complexity of the symbol-wise
Householder QRD, the BH-QRD can save about 75\% of the computational complexity. We also propose the block-wise complex Givens rotation based QR
decomposition (BCGR-QRD) that can save about 57\% of the number of rotation operations compared with that of the symbol-wise Givens QRD. Besides, the
block-wise QRDs are extended to the block-wise sorted QRD that is an ordering QRD at the preprocessing stage of the detection.

For the purpose of further performance improvement, we apply the Tomlinson -Harashima precoding (THP) technique to the layered Alamouti STBC system,
and present three precoding schemes, including group-wise MMSE THP, group-wise LMSE MMSE THP, and real-valued LMSE MMSE THP. The simulation results
show that the developed precoded system can provide superior performance compared with the non-precoded layered Alamouti STBC system.

We also study the symbol error rate (SER) analysis for the zero-forcing detection with successive interference cancellation (ZF-SIC). Relying on the
results obtained from the matrix variate distribution, we evaluate the exact SER of the ZF and ZF-SIC detection for the layered Alamouti STBC system
under the independent and identically distributed Rayleigh fading channel. These analytical SER curves are shown to agree with the SER curves from
Monte Carlo simulations.

並列關鍵字

Tom linson -Harashima precoding ； symbol error rate ； QR decomposition ； block-wise complex Givens rotation ； block-wise Householder based QR decomposition ； layered Alamouti STBC ； space-time block coding (STBC)

參考文獻

[30] P. Y. Liu, Hardware Implementation of High-Throughput QR Decomposition for

an architecture for realizing very high data rates over the richscattering wireless

[2] V. Tarokh, H. Jafarkhani, and A. R. Calderbank, “Space-time block code from

orthogonal designs,” IEEE Trans. Inform. Theory, vol. 45, pp. 1456-1467, July

[4] X. Zhu and R. D. Murch, “Performance analysis of maximum likelihood detec-