由於良好的頻譜使用效率以及對抗多路徑通道的能力，OFDM 系統已成為下一代無線通訊系統的主要技術。在多使用者的環境中，多存取 OFDM 系統或稱之為OFDMA成為一個重要的技術並且開始被重視。然而由於不同的使用者可能來自不同的行動通道環境，加上使用者端頻率震盪器先天的誤差，導致多用戶的 OFDMA 系統面臨比單用戶系統裡更嚴格的同步要求。在本計畫中，為解決 OFDMA 在未來無線通訊上的問題，針對多載波頻率漂移問題以及高速行動通訊造成的 ICI 效應進行研究。我們提出新的多重 CFOs 補償方法，以獲得 MMSE 條件下的最佳性能，並且透過兩年期的計畫，除了以數值解法實現MMSE 的目標之外，我們可以發展出理論結果並作 SIR 性能分析的研究，以驗證先前提出的數值方法。在另一方面，我們針對行動通訊所造成的 ICI 效應提出新的決策迴授干擾消除法，搭配先前已發表的通道估計誤差補償方法，以及利用時域 Kalman filter 做通道脈衝響應估計，構成一個全新的 DFIC 演算法。在兩年期計畫的安排上，我們將能完整探討演算法、系統參數、以及低複雜度替代方案，本計畫將能對國內在 4G 技術的研究與人才培育上提供相當重要的貢獻。Orthogonal frequency division multiplexing (OFDM) becomes the main technology for the next generation of wireless communication system since it has high spectral efficiency and is resistant to multipath channels. In a multiuser environment, the multi-access OFDM (OFDMA) system is so important that many interests have been mentioned. Due to different subscribers coming from independent mobile channel environments and the mismatch of local oscillators at transmitters, the requirement of synchronization for an OFDMA system is more stringent than that for a single-user OFDM system. In this project, for the purpose of solving possible problems to be faced in future wireless communications, we propose the research to deal with multiple carrier frequency offsets (CFOs) and the inter-carrier interference (ICI) effect due to high mobility. We propose a new multiple CFOs compensation method in order to obtain the minimum mean square error (MMSE) performance. With the help of two years of research grant under request, we can develop theoretical results with the analysis of signal-to-interference (SIR) index in addition to the already developed numerical implementation method for approaching MMSE. On the other hand, we also propose a new decision feedback ICI cancellation (DFIC) method for mobile communications. In cooperation with the previously published channel estimation error concealment method and the time-domain Kalman filter for channel impulse response estimation, we can construct a new and novel DFIC algorithm. In our two years of project schedule, we can thoroughly study algorithms, system parameters, and the alternative of low complexity. This project can definitely contribute to research and talented people related to the 4G technology in Taiwan. 研究期間：10008 ~ 10107
