上行鏈路單輸入多輸出單載波分頻多重存取系統的迭代決策回授混合式等化方法

Abstract

在本論文中，我們提出將迭代式決策回授混合等化（iterative decision-feedback hybrid equalization; IDFHE）方法結合後等化等增合併（equal-gain combining; EGC），以提升上行單輸入多輸出（single-input multiple-output; SIMO）單載波分頻多重存取（single-carrier frequency-division multiple-access; SC-FDMA）系統於多徑通道下的符號判決效能。IDFHE 透過在不同迭代階段分別採用決策回授等化與混合等化器，得以逐步抑制等化所造成的殘餘符號間干擾（inter-symbol interference; ISI），並於最後一次迭代避免雜訊放大。在SIMO架構下，每次迭代結束後皆整合 EGC 於符號判決，以充分發揮空間多樣性。模擬結果顯示，所提出的 SIMO IDFHE方法在錯誤效能及對載波頻率偏移（carrier frequency offset; CFO）與通道估測誤差（channel estimation error; CEE）之強健性方面，均顯著優於傳統 SIMO等化方法。此外，SIMO IDFHE方法也可應用於加入前向錯誤修正碼(forward error correction; FEC）的 SC-FDMA 系統中，進而顯著降低錯誤率。

In the thesis, an iterative decision-feedback hybrid equalization (IDFHE) method is integrated with post-equalization equal-gain combining (EGC) to enhance equalized symbol decision over multipath channels in uplink single-input multiple-output (SIMO) single-carrier frequency-division multiple-access (SC-FDMA) systems. The IDFHE method progressively mitigates residual inter-symbol interference caused by equalization by combining decision-feedback ISI excision with hybrid equalizers across multiple iterations, and avoids noise amplification by applying a constrained equalizer in the final iteration. With SIMO reception, post-equalization EGC is integrated with equalized symbol decision at each iteration to exploit spatial diversity. By simulation, the SIMO IDFHE method is shown to outperform conventional SIMO methods remarkably in error performance and robustness against carrier frequency offsets and channel estimation errors. In addition, the SIMO IDFHE method can be applied to FEC-coded SC-FDMA systems, resulting in a noticeable decrease in error rate.