多入/多出通訊系統之新穎傳送與接收技術

Abstract

多入/多出(multiple-input, multiple-output, MIMO)是適用於多種通訊系統之系統模型，含多傳送/接收天線系統(multiple transmit and receive antenna systems)、分碼多重存取 (code-division multiple-access systems, CDMA)系統及符元間相互干擾系統(systems with inter-symbol interference channels)等。本論文針對MIMO系統提出新型傳送及接收處理技術。首先，本論文針對空間多工(spatial multiplexing)之多天線系統，提出兩種新型MMSE(minimum mean-squared error)通道分解協助式偵測器(channel-factorization aided detector)；從硬體複雜度(hardware complexity)或位元錯誤率(bit-error-rate)上作比較，新型的偵測器具優於現存方法之效能。接著，本論文針對多天線MIMO-OFDM (orthogonal frequency-division multiplexing)系統在I-Q失衡的干擾下，研究接收機架構並在運算複雜度及位元錯誤率的取捨下作最佳的設計。最後，本論文則針對二維 OFDM-CDMA下行 (forward link)系統提出新型傳輸端預處理技術(transmitter-based pre-processing)；數值模擬結果顯示： (i) 根據可達到之總資料速率，所提出的新方法之效能優於其他現存方法，(ii) 如果能夠依據通道選擇特性(channel selectivity)來適當的決定延展樣式(spreading pattern)，則二維延展(2-D spreading)之效能優於一維延展。

Multiple-input, multiple-output (MIMO) is a model for a variety of communication systems, including the multiple transmit and receive antenna systems, code-division multiple-access (CDMA) systems, systems with an inter-symbol interference channel, etc. In this dissertation, new transmit and receive techniques for the MIMO communication systems are proposed. Firstly, two new MMSE (minimum mean-squared error) channel-factorization aided detectors (CFAD) are proposed for the spatial multiplexing multiple antenna systems. The proposed detectors outperform the existing ones in terms of hardware complexity and/or bit error rate performance. Secondly, receiver designs with good trade-off between computational complexity and BER performance are investigated for the MIMO-OFDM (orthogonal frequency-division multiplexing) systems with the presence of I-Q imbalances. Lastly, new transmitter preprocessing techniques are proposed for the two-dimensional (2-D) OFDM-CDMA forward link systems. Numerical results show that (i) the new method outperforms the existing ones in terms of achievable sum data rate, and (ii) the 2-D spreading outperforms the 1-D one if spreading pattern is adapted according to channel selectivity.