基於正交分頻多工系統中雙選擇性通道週期性訓練序列之到達角，都卜勒頻率偏移及通道之估測

Abstract

在無線通訊系統中，因為正交分頻多工在頻率選擇性通道中具有優勢，故常為人使用。然而，在多數移動應用中，由於發射器或接收器的快速移動，通道變得具有時域選擇性，這將導致正交分頻多工系統的性能下降。這種同時具有頻率和時域選擇性的通道被稱為雙選擇性通道。近年來，關於正交分頻多工系統中雙選擇性通道的研究變得更為熱門。在這篇論文中，我們採用了一個基於分接延遲線的雙選擇性通道模型來研究都卜勒效應在正交分頻多工系統中的影響。

在許多通訊系統中，傳送端會傳送一個週期性的前導信號。在這篇論文中，我們將會展示如何利用前導信號的週期性質來估計都卜勒頻率偏移和通道。與之前的方法不同，我們的方法能夠在一根天線的情況下進行都卜勒頻率偏移的估測。我們利用了旋轉不變技術的信號參數方法以及最大似然方法來估計都卜勒頻率偏移。此外，我們得出了克拉瑪下界，並且模擬結果顯示我們提出的都卜勒頻率偏移估計器的均方誤差接近於克拉瑪下界。

我們也將單天線的情況延伸到均勻線性陣列。與單天線的情況類似，週期性訓練序列可以被用於到達角、都卜勒頻率偏移和通道的估測。在這種情況下，二維的旋轉不變技術的信號參數估計方法可以被用來估計到達角和都卜勒頻率偏移。模擬結果表明，我們所提出的方法在天線增加時，可以得到更好的效能。

In wireless communication systems, orthogonal frequency-division multiplexing (OFDM) is one of the popular schemes since it has an advantage in frequency-selective channels. However, in many mobile applications, the channels become time selectivity because of the fast-moving transmitter or receiver, and this will cause performance degradation to OFDM systems. Channels with both frequency and time selectivity are called doubly selective channels. The study on doubly selective channels for OFDM systems has become a popular research topic in the last few years. In this thesis, a tapped delay line (TDL) model of doubly selective channels is adopted for the study of the Doppler effect in OFDM systems.

In many practical communications, the transmitter often sends a preamble consisting of periodic training sequence. In this thesis, we show how the periodicity can be exploited for the estimation of Doppler frequency offsets (DFOs) and channel. Unlike the existing methods, our method is able to perform the DFO estimation with a single received antenna. Both unitary estimation of signal parameters via rotational invariance techniques (ESPRIT) method and maximum likelihood (ML) method for DFO estimation will be derived. Moreover, the Cramér–Rao bound (CRB) is also derived and simulation results show that the mean squared errors (MSE) of the proposed DFO estimators are close to the CRB.

We also extend our method to the case of uniform linear array (ULA) antennas. Similar to the single antenna case, the periodic training sequence can be exploited for the joint estimation of angle of arrival (AoA), DFO, and channel. In this case, the two-dimension unitary ESPRIT can be used to estimate AoAs and DFOs. Simulation results show that the performance of the proposed system improves as the number of antennas increases.