近年來由於無線區域網路的蓬勃發展,使用多輸入多輸出 (MIMO) 系統已經成為無線區域網路發展的主流趨勢。因此衍伸出許多相關技術。其中有限回饋預編碼選擇技術是本篇論文探討的重點。在有限回饋預編碼系統中可以使用編碼簿的設計,並利用挑選準則從中挑出最合適的碼字作為預編碼矩陣。在本篇論文中我們使用 Kerdock 編碼簿並根據傳送串流數來搜尋碼字的集合。儘管錯誤率效能最佳的搜尋方法為全搜索演算法,但以此搜尋碼字所耗費的計算複雜度過大,因此本篇論文提出三種碼字搜索演算法來降低計算複雜度。這三種演算法分別為碼字遞增搜索演算法、碼字遞增搜索快速演算法、以及串流配對碼字搜索演算法。 由於 Kerdock 編碼簿之碼字不論傳送串流數皆從同一組編碼簿中做搜尋。因此能進一步的使用在多重模式切換有限回饋預編碼選擇技術。此技術為一種結合了傳送端訊號個數選擇以及有限回饋預編碼選擇的技術,並且可以降低空間多工 MIMO 系統之錯誤率。而在挑選準則上我們以使最大 MSE 值最小化當作切換準則。並且我們利用此技術結合接收端連續干擾抵消 (successive interference cancellation, SIC) 或線性偵測的架構。最後透過電腦模擬結果來顯示不同系統其錯誤率效能以及複雜度比較。
In recent years, multiple-input-multiple-output (MIMO) communications have become a key technology due to the increasing demand of high speed wireless communications services. Among a number of MIMO techniques, limited feedback MIMO precoding has drawn increasingly interests as it can provide substantial closed-loop gain at the expense of reasonable feedback traffic overheads. In this thesis, we investigate a number of codeword selection schemes for different transceiver architectures using Kerdock codebook. In contrast to other limited feedback precoding codebooks such as the Grassmann codebook, the Kerdock codebook has the capability of supporting multiple precoding modes with the same codebook. As a result, The Kerdock codebook has the advantage of offering low storage complexity and low feedback overhead when applied in multi-mode precoding. We apply the proposed codeword selection schemes to both linear and successive-interference-cancellation transceiver design, and compare their performance, required complexity, and amount of feedback information via extensive computer simulations.