高斯-馬可夫衰減通道下之編碼結構

Abstract

2002年，史克蘭(Skoglund) ，蓋斯(Giese)和巴克孚(Parkvall)介紹了一個結合通道估計(channel estimation)，效應等化(equalization)和解碼(decoding)的觀念，他們設計一可同時提供接收端通道估計與錯誤保護的區塊碼使區塊錯誤率達到最佳化，然而在極快速變動的通道環境中他們的方法是否同樣適用並未獲得證實。在本篇論文中，我們將其方法擴展為能適用在高斯-馬可夫通道(Gauss-Markov channels)下，以推導出的成對錯誤發生率(pairwise error probability)作為電腦模擬鍛鍊演算法(simulated annealing algorithm)的判斷準則，相對於[7]的結果，當區塊錯誤機率(WER)為0.01且高斯-馬可夫通道長度1與長度2的情況下，我們的設計碼得到約3.5dB與6dB的改善。

In 2002, Skoglund, Giese and Parkvall introduced a novel concept of combining channel estimation, equalization and decoding. They optimized the block error rate by designing a block code which can provide channel estimation and error protection to the receiver at the same time. However it is not clear that their approach will perform well on a fast-varying channel. In this thesis, We extend their approach in order to adjust well to Gauss-Markov channels. By deriving the pairwise error probability as a function of the criterion for the simulated annealing algorithm, Our designed codes provide a coding gain of about 3.5 dB and 6 dB on the Gauss-Markov channels with channel lengths 1 and 2 respectively over those given in [7] at WER= 10¡2.