Channel Estimation Techniques with Various Pilot Assignments for Wireless STBC-Based MIMO-OFDM Communications

[[abstract]]由於傳收技術的快速發展，未來無線通訊系統必須兼具高傳輸速率、有效率的利用頻寬、以及提高傳送訊號之品質。對於正交分頻多工 (OFDM) 傳輸系統而言，能夠極為有效地來降低多重路徑干擾下所產生的頻率選擇性衰落效應，同時頻寬資源的使用效率亦可以獲得提昇，是為其最大的特點。多輸入多輸出 (MIMO) 之傳輸架構能夠提供多個獨立的無線通道，透過裝置於傳送與接收兩端上的多根天線來同時收送多個訊號，運用空間分集的技術提昇訊號品質。將多天線之傳輸結構與正交分頻多工系統相互結合之後，即可形成所謂的多輸入多輸出正交分頻多工 (MIMO-OFDM) 無線傳輸系統，預期可以同時兼具前述兩種傳輸機制的優點，使得通訊的品質能夠更進一步地獲得提昇。而在近年來，領航信號輔助的通道估測技術應用於正交分頻多工系統中已經相當地成熟，許多研究成果也驗證了其良好的估測效果與便於實現的特色。然而此類通道估測技術若運用於多天線架構下的正交分頻多工系統時，領航信號之設計與配置上的調整則是必需的，同時其估測結果將直接影響到多天線系統的接收端對於多重接收訊號之時空區塊解碼的正確度。在本論文中，在多根傳送天線之架構下設計領航信號，最關鍵的性質便是領航信號彼此間必須相互保持正交，使得接收端可以利用此正交性正確地完成多重通道之估測，而領航信號將被配置成梳形樣式 (comb-type) 以適用於頻率平坦性衰落快速時變通道。在此我們提出了新的通道估測演算法，搭配正交梳形配置的領航信號，應用於時空區塊碼之二輸入二輸出正交分頻多工（STBC-based 2I2O-OFDM）系統在多重路徑為四條路徑鄉村（4-path RA）、六條路徑一般城市（6-path TU）與六條路徑惡劣城市（6-path BU）環境中來完成雙重通道之估測。本論文所採用之通道估測演算法分別如下：（1）內積式（inner product）結合非零領航信號（non-zero pilot）之通道估測，（2）內積式結合虛擬領航信號（virtual pilot）之通道估測以及（3）最小平方（least squares）結合虛擬領航信號之通道估測。以上所述的方法將透過一些簡易的數學分析和模擬圖在4-path RA、6-path TU與 6-path BU環境中做分析以及歸納結果比較，詳實地探討領航信號輔助式通道估測技術應用於時空區塊編碼之多輸入多輸出正交分頻多工系統，在多種戶外無線通訊環境中之效能，並且說明領航信號在多天線系統中的設計特點與配置方式，將其優點與便利性加以推廣實現，以達成整體系統效能與通訊品質均獲得增進與保障之最終目標。[[abstract]]Due to the rapid development of transceiving techniques, high transmission data rate, efficient spectral bandwidth, and quality of reliability are more necessary for future wireless communications systems. Orthogonal frequency division multiplexing (OFDM) schemes are a well-known wireless transmission technique which is effective against frequency selective fading channel and capable of reducing multipath propagation effects very well. Multi-input multi-output (MIMO) architectures can provide multiple artificial uncorrelated channels for transmitting/receiving replicas of the same signals simultaneously through various antennas at different symbol duration to obtain diversity gains. For MIMO-OFDM systems, channel estimation is crucial for the decoding procedure of space-time block coding (STBC) to separate the superposition results from multiple transmitted signals. In order to compensate the signal distortion, pilot-symbol-assisted channel estimation (PSACE) techniques with various pilot assignments have been used to obtain the channel impulse response (CIR). In practice, the channel estimation procedure can be assisted by transmitting pilot symbols that are already known at the receiver end. Therefore, system performance with channel estimation usually depends on the choice of pilot patterns and the number of pilot symbols. In this thesis, three kinds of channel estimation algorithms are developed with various pilot assignments for STBC-based MIMO-OFDM systems. They are (1) inner product estimation with non-zero pilots (M1), (2) inner product estimation with virtual pilots (M2) and (3) least squares estimation with virtual pilots (M3). All of three estimation algorithms utilize orthogonal pilot patterns between Tx.1 and Tx.2. The performances of M2 and M3 are almost identical. If there are more antennas on transmitter end, M3 is easier to implement than M1. Therefore, the estimation algorithm in M3 is not only to reduce computation complexity but also to obtain better system performance. Finally, some comparative simulations are given to illustrate the superior performance of M3 in STBC-based MIMO-OFDM systems under COST 207 mobile wireless environments.[[note]]碩