多天線頻域正交多工系統之頻率,資料與通道的合併估測

Abstract

我們探討在頻域正交多工(OFDM)系統中合併頻率、資料與通道的估測的可行性。由於載波頻率偏移(Carrier Frequency Offset, CFO)所造成的頻域信號的載波間干擾(Inter-Carrier Interference, ICI)與複徑通道所引起在時間軸上符碼間干擾(Inter-Symbol Interference, ISI)有類似的數學模式，後者的許多已知解決方案，如最大可能性序列估測(Maximum Likelihood Sequence Estimation, MLSE)，便可用來解決前者。然而，不同於傳統的MLSE可直接套用Viterbi演算法，我們面對的是含有未知參數（不確定性）的籬柵(Trellis)圖。我們採用遞迴式估合併檢測與估計(Iterative Joint Detection and Estimation)的原則來解決。

首先，我們針對單天線 (SISO) 系統的狀況來設計，並且針對效能及複雜度等主要課題提出改進方案。我們討論了利用接收端視窗(Windowing)的等效濾波及刪除較小載波干擾的鄰頻等降低籬柵狀態(trellis state)數目的方法的可行性以降低演算法的複雜度。我們也探討採用不同通道估測法對系統效能的影響。接著，我們將這套遞迴式估合併檢測與估計演算法推展到多天線(MIMO)的系統。

為研究我們所提出演算法的效能及各系統及通道參數的效益，我們採取IEEE802.11n標準，Task Group‘n’synchronization（或簡稱TGn Sync）所提出來的傳輸信號架構為基礎來做電腦模擬。所得到的數值結果顯示，無論是SISO或MIMO環境，我們的解決方案都能提供令人滿意的表現。

We consider the problem of joint carrier frequency ofset (CFO)/channel estimation and data detection for orthogonal frequency-division multiplexing (OFDM) systems. An iterative approach is employed. The iteration procedure consists of an inner CFO/data estimation loop and an outer CFO/data-channel estimation loop. We first discuss the single-input single-output (SISO) case and then extend to a multiple-input multiple output (MIMO) scenario. Drawing an analogy between the inter-carrier interference corrupted frequency domain waveform and the inter-symbol interference limited time domain waveform, we apply maximum likelihood sequence estimation scheme to detect the data sequence in the presence of residual CFO. Related algorithms either for reducing the implementation complexity like the trellis state number or for improving the overall performance such as receiver time domain windowing and re?ned channel estimates are presented. Computer simulation results based on a candidate IEEE 802.11n standard-the Tgn Sync (Task Group n) proposal-are given to predict the performance of the proposed algorithms.