IEEE 802.16m 初始測距探討

Abstract

本篇論文研究方向為針對IEEE 802.16m 中初始測距程序 (initial ranging process)探討，我們探討了時間位移 (timing offset)，以及頻率估計演算法，同時我們亦分析遺失偵測機率 (miss detection rate)和假警報機率 (false alarm rate)在加成性白色高斯雜訊 (additive white Gaussian noise, 簡稱AWGN)通道下的理論值。 當一個行動電話開始要進入網路的時候，我們必須與基地台做下行同步 (downlink synchronization)去取得相關的參數去做上行同步，也就是初始測距程序。在初始測距程序，基地台會對所有接收到的測距碼做估測，包含了時間偏移、載波偏移和測距碼 (ranging code)需要同步估測。然後，基地台以廣播的方式將每組接收到的測距碼以及其所對應的時間和頻率調整資訊回傳。此外，測距的成功與否也已廣播的方式告知。 在頻域上，我們大量的遞減時間位移估計所需的乘法運算量，而與時域上所 用互相關演算法比較，我們所使用的演算方式只需要0.5%的運算量。我們也從理 論的觀點分析遺失偵測率和假警報率在加成性白色高斯雜訊通道下的理論值 ， 跟電腦模擬的結果比起來，我們分析的比較保守。給定遺失偵測率和假警報率不 同權重，我們可以決定基地台的門檻值 (threshold value)去判定測距使用者 (ranging user)存在與否。 最後，我們使用重建訊號 (regenerated signal)和疊代技巧，以及測距訊號 (ranging signal)，我們提出了一個架構去做小數載波頻率偏差 (fractional carrier frequency offset，簡稱FCFO)偵測，改善了原本方法的準確度，即使在多路徑通 道 (multipath channel)都仍有不錯的效能。

In this thesis, the research focus on initial ranging process of IEEE 802.16m, and discuss the algorithm of timing offset and frequency estimation. We also analyze theoretical miss detection and false alarm rate in the AWGN (additive white Gaussian noise) channel. When a mobile station enters the network, it needs to perform initial downlink synchronization to get related parameters for use of uplink synchronization, so-called “ranging process.” In the ranging process, BS (base station) is required to detect different received ranging codes, including symbol timing offset, carrier frequency offset and ranging code. Then BS broadcasts the detected ranging codes with adjustment information of the timing and frequency. We reduce the amount of multiplication computation for timing offset estimation in frequency domain. Comparing to cross correlation computation in timing domain, our method need 0.5% computational amount of original method. We also analyze theoretical miss detection and false alarm rate in the AWGN channel, comparing to computational simulation results, our analysis is a bit conservative. Given miss detection and false alarm rate different weights, we can decide BS’s threshold value to judge whether ranging user exists or not. Finally, we propose a structure to estimate FCFO (fractional carrier frequency offset) by using techniques of regenerated signal and iteration and specific ranging signal format. It improves the performance compared to the original method in the AWGN channel. Even in the multipath channel, it has better performance.