頻率選擇性衰減通道中對於外層渦輪碼正交頻分複用系統之內層迴旋碼設計

Abstract

在一個未知通道係數的頻率選擇性衰減通道中，通道估測以及通道等化基本上難以實現。因此在此篇論文，基於碼的累計距離函數(cumulative distance function)，我們提出一個簡單的法則來選擇內層迴旋碼，以與外層渦輪碼共同合作進行資料傳輸保護。在我們的設計中，內層迴旋碼其實相當於一個通道估測/等化器。由於我們所設計的內層迴旋碼的碼長極短，因此在接收端可以使用基於一般最大概度檢定測試準則(GLRT)的窮盡解碼法。模擬結果顯示，我們的方法在碼率為1/2時可以得到相當好的系統效能，但是當碼率提昇至2/3時，則所選出的迴旋碼有時仍非最佳選擇。我們最後經由模擬比較，確認使用我們所設計的內層迴旋碼確實可以達到比傳統最小平方(least square)通道估測更好的系統效能，故而我們的設計應該可以作為全盲通訊環境下，一個相當好的系統設計選項。

In frequency-selective fading channels with unknown channel coefficients, channel estimation and equalization become infeasible. In this thesis, based on the cumulative distance functions of codes, we propose a simple method to select an inner convolutional code to co-work with an outer turbo code, which acts implicitly as a channel estimator and equalizer. Since the codeword length of the selected convolutional code is short, an exhaustive decoder based on the GLRT criterion can be implemented at the receiver end. Simulation results show that the concatenated coding system that use the convolutional codes of rate 1/2 performs well but may have rooms for improvement when the inner code rate increase to 2/3. In comparison with the tradition system that employs a least square (LS) estimator for the outer turbo code, our proposal can result in better performances and hence can be a good candidate design in a blind frequency-selective fading environment.