標題: 多傳送多接收位元交錯調變碼系統之低複雜度迭代訊號偵測設計
On the Design of Low Complexity Iterative Signal Detection for MIMO BICM Systems
作者: 曾鼎哲
Ting-Che Tseng
沈文和
Wern-Ho Sheen
電信工程研究所
關鍵字: 多傳送多接收;位元交錯調變碼;正交分頻多工;最小均值平方差;迭代訊號偵測;渦輪原理;MIMO;BICM;OFDM;MMSE;Iterative Signal Detection;Turbo principle
公開日期: 2004
摘要: 隨著數位多媒體時代的來臨,用戶對於資料的需求急速增加。下一代的無線通訊系統,如無線區域網路(802.11n)、第四代行動通訊系統,將可能採用多根天線傳送及接收(MIMO)技術以提高資料傳輸率。然而,要如何在現有的通訊系統中實現此新的技術是近年來熱門的研究。現行的通訊系統使用正交分頻多工(OFDM)技術和位元交錯調變碼(BICM)來克服多重路徑、瑞雷(Rayleigh)衰落通道,以提升系統效能。因此,此篇論文主要是在下一代通訊系統中,設計低複雜度迭代訊號偵測。在低複雜度零強制(ZF)和最小均值平方差(MMSE)訊號偵測器中,利用近似方法推導位元度規(bit metrics)的計算。另外,藉由渦輪(Turbo)原理,設計低複雜度迭代MMSE偵測器,並提出幾個近似的方法減少偵測器的運算。最後,採用下一代無線區域網路802.11n提案的系統架構,作為系統模擬環境。利用電腦模擬方式,印證使用近似的位元規度計算,能有效地提高系統效能。此外,在迭代MMSE偵測器中,從模擬結果顯示,利用這些近似的方法能降低計算的複雜度,但不會減弱系統效能。
With the advent of digital multimedia communications era, the amount of the demand for data stream of subscribes is increasing rapidly. The next generation of wireless communications, such as 802.11n wireless local area networks (WLAN), 4G mobile communications, may utilize multiple input multiple output (MIMO) approach to enhance data rate. Existing communications use orthogonal frequency division multiplexing (OFDM) and bit-interleaved coded modulation (BICM) techniques to overcome multipath Rayleigh fading channels. Hence, the theme of my thesis is to design low complexity iterative signal detection for next generation of wireless communications. We derive the bit metrics based on zero-forcing (ZF) and minimum mean squared error (MMSE) detector by approximation. Besides, we design low complexity iterative MMSE detector with turbo principle and propose some methods of approximation to reduce computation complexity. Finally, we apply them to the system model of 802.11n Proposal. From simulation results, it proves that using approximated bit metrics can improve the performance, and employing the approximation of iterative MMSE detector can reduce the computation complexity without performance deterioration.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT009213502
http://hdl.handle.net/11536/69423
Appears in Collections:Thesis


Files in This Item:

  1. 350201.pdf