IEEE 802.16m 線性最小均方差通道估測之數位訊號處理器實現

Abstract

本篇論文介紹IEEE 802.16m 正交分頻多工存取(OFDMA)於上行傳輸以及 下行傳輸的通道估測部分，並採用空間頻率區塊碼(SFBC)以對抗抗通道衰減， 之後再將其引入DSP 定點運算並觀察分析其效能。 本篇論文採用線性最小均方差的通道估測技術。首先我們先使用最小平方差 的估測器估計在導訊(pilot)上的通道響應，再利用導訊上的通道響應使用線性內插法得到其餘導訊位置上的通道響應。之後估測延遲參數，再利用指數函數的功率延遲曲線求出相關函數，最後利用相關函數作線性最小均方差估測資料載波上的通道響應。 本篇論文可大致分為三個部份。第一部份我們先簡介IEEE 802.16m 於上下 行傳輸的相關規格制定。第二部份我們介紹最小均方差演算法並驗證其採用浮點 運算於AWGN 通道以及多重路徑SUI 通道上的效能。第三部分我們介紹DSP 晶 片實現環境並修改浮點運算程式碼為定點運算，然後針對定點運算程式碼進行優 化及加速的動作。最後再分析其在AWGN 及多重路徑SUI 通道上的效能。

In this thesis, we discuss about IEEE 802.16m OFDMA uplink and downlink channel estimation and adopt SFBC (special frequency block code) technique to resistant the channel fading. After that we use DSP fixed-point operation to analysis the performance. We present LMMSE channel estimation method in this thesis. First, we use LS estimator on pilot subcarrier to get the channel response. And then we use linear interpolation method to get channel response on other pilot position. After that we estimate the delay parameters and find correlation function associate exponential PDP. Finally base on this correlation function, do LMMSE filtering to estimate the data subcarrier response. This thesis can be separated into three parts. First we introduce about the standard of IEEE 802.16m OFDMA uplink and downlink transmission. Second, we introduce LMMSE algorithm and validate on AWGN and SUI multipath channel to analysis the performance on floating-point condition. Third we introduce DSP implementation environment and modify the floating-point C code to fixed-point. And then we against the fixed-point C code to accelerate and optimization. Finally we do the simulation on AWGN and SUI multipath channel to analysis the performance.