OFDM系統可提供高速傳輸的調變與解調變技術,應用於數位行動通訊系統及高密度之儲存裝置上,不過由於傳送端與接收端的振盪器無法產生很穩定的頻率,因而會伴隨產生相位雜訊,而這相位雜訊是時變且對每個正交分頻多工系統符元會產生兩種影響,一種是共同相位錯誤(Common Phase Error, CPE),另一種則是子載波間干擾效應(Inter-Carrier Interference, ICI)。 本論文應用可變步階最小均方(Step Size Delta-LMS, SSD-LMS)演算法做相位誤差消除,SSD-LMS演算法與傳統LMS演算法比較,可減少疊代次數與位元錯誤率(BER),而模糊步階最小均方(Fuzzy step size delta-LMS, FSSD-LMS)演算法跟SSD-LMS演算法比較起來,更可快速收歛和降低位元錯誤率,模擬結果顯示我們所提出改良的LMS演算法具快速收斂和追蹤的特性。
Orthogonal Frequency Division Multiplexing (OFDM) is a significant technology to provide a high-rate wireless transmissions. However, with applying OFDM technology, one of the principal disadvantages of OFDM is its sensitivity to the common phase error (CPE) and inter-carrier interference (ICI) caused by the random Wiener phase noise. In this paper, a step size delta - least mean square, (SSD-LMS) is used for phase error cancellation. In our scheme, SSD-LMS compared to convention LMS algorithm can reduce the number of iterations and bit error rate (BER), while the fuzzy step size delta - LMS, (FSSD-LMS) with the SSD-LMS in comparison, but also fast convergence and smaller bit error rates. Therefore, the results show that the proposed algorithm can provide a rapid convergence with a low steady-state fluctuation error than that with the conventional LMS algorithm.