新式Viterbi演算法及802.11a編解碼器設計

Abstract

本論文為高階數二維振幅調變(QAM)以及以位元為單位交錯編排(interleave)的廻旋碼(convolutional code)，提出了一個新的”新式Viterbi高精度(soft decision)解碼演算法”。這個新的演算法可以明顯的改進編碼增益。例如在IEEE 802.11a規格下，所提出的演算法可以比低精度(hard decision)多出3.3dB的增益，比傳統的高精度演算法多出1dB的增益。另外這篇論文也提出了針對802.11a通道編碼的電路設計。目前最新的無線區域網路的規格都集中於發展高資料量的傳輸速率，而其中之一的瓶頸就是實現一個高速率的通道解碼電路，因為在無線區域網路中的解碼速度是目前行動裝置解碼速度的一千倍。從量測結果顯示，經由一些特殊的設計方式，我們所設計出的Viterbi解碼器其解碼速率可以達到100Bits/s，而規格所要求的最高速率是54Bits/s。

In this thesis, a modified Viterbi soft-decision decoding algorithm for high QAM and bit-base interleaving is presented. This modified algorithm can significantly improve the coding gain. For example, in IEEE 802.11a the proposed algorithm has 3.3dB over hard decision and 1dB over traditional soft decision in 64QAM and 3/4 code rate. Also the implementation of Viterbi decoder in IEEE 802.11a has been developed. The latest standards of WLAN are focused on the high data throughput. One of the bottlenecks is to implement a high speed channel decoder where the decoding speed requirement must be 1000 times faster than vehicular application. The measurement result has shown that, through proper design, the Viterbi decoder in FPGA can achieve 100Mbits/s while the highest data rate in IEEE 802.11a is 54Mbits/s.