在向上鍵結正交分頻多重接取系統中，載波頻率偏移將會引進載波間干擾(ICI)和多重存取干擾(MAI)，將會降低系統效能。盲目載波估計演算法中以[1]所提出的多重信號特徵(MUSIC)受到相當關注，但這個演算法在硬體時作上有兩個高複雜度的運算須克服，分別是雜訊子空間的獲得及依解析度大量蒐尋估測參數，此外在無線傳輸單位是一個偵框，而通道和在波頻率偏移應該是隨著每一個符元變化，在每一個符元時間下做一次估計是沒有效率的，在此篇論文之中，我們引用了快速資料投影法(FDPM)來獲得雜訊子空間，並設計了一個迭代多重信號特徵演算法以快速完成一個偵框長度的載波頻率偏移的估測。In OFDMA uplink systems, carrier frequency offset (CFO) would cause inter carrier interference (ICI) and multiple access interference (MAI), and would degrade the system performance. Blind CFO estimation algorithms have been proposed in lot of papers, and [1] proposed the MUSIC algorithm with excellent performance, but the MUSIC algorithm includes two operation that are hard for hardware to implement, to acquire the noise subspace and grid search procedure that depends on search resolution, besides, the wireless communication system transmission unit is frame, but the channel and CFO will change along the symbols, so if we estimate the CFO per symbol isn’t efficiency, so in this paper, we applied a fast data projection method (FDPM) to get the noise subspace and proposed a iterative MUSIC scheme to estimate the CFO over frame rapidly.
