時空正交分頻多工調變之盲式通道估計

Abstract

時空正交分頻多工調變不僅可以達到高傳輸效率而且可以利用分集增益(diversity gain)，所以目前倍受矚目。在本篇論文中，我們採用了 Giannakis 所提出的時空正交分頻多工調變之盲式通道估計，並且為此估計子推導出均方誤差。不僅如此，我們更介紹了 decision direct (DD) 和 phase direct (PD) 兩個方法來使之前所提的通道估計更加趨於理想。 DD 是利用解得的決策信號來更新通道的估測，而 PD 是在我們得到通道的振幅響應之後，解得其相角響應，原來是只應用在一般的正交分頻多工上， 因為其通道的振幅響應很容易得到，但是在時空正交分頻多工調變上，因為接收到的信號是由兩個傳送天線傳來的，通道的振幅響應很難由其中解出。所以我們提出了一和差平方的演算法來的到通道的振幅響應進而就可以用 PD 來使通道的估計更趨於理想。並且在電腦模擬中驗證了我們的方法真的使通道的估計更完美。

Space time (ST) orthogonal frequency division multiplexing (OFDM) has been well documented as an attractive means of achieving high data rate transmissions with diversity gains. In this thesis, we adopt a blind channel estimation algorithm proposed by Giannakis for ST OFDM, and derive the theoretical mean square error of the estimator. Moreover, we introduce phase direct (PD) and decision direct (DD) methods to further improve the performance of the estimator. DD and PD originally work on conventional OFDM, and PD is not suited for ST OFDM. Then we derive a new algorithm named sum-difference square method to make PD work on ST OFDM. DD is to update our estimated channel from the previous hard decision data, while PD is to solve the phase ambiguities after we’ve got the channel power response. Since the received data in ST OFDM is composed of two different transmitted data, the channel amplitude response is not easy to get. Hence, the aforementioned algorithm is about how to solve this problem. Furthermore, in computer simulations, we can see our algorithm really better the performance.