ML Estimation of Timing and Frequency Offsets Using Distinctive Correlation Characteristics of OFDM Signals Over Dispersive Fading Channels
ML Estimation of Timing and Frequency Offsets Using Distinctive Correlation Characteristics of OFDM Signals Over Dispersive Fading Channels
复制标题
DOI：
10.1109/tvt.2010.2102058
复制
发表时间：
2011-02
影响因子：
6.8
通讯作者：
Wen-Long Chin
中科院分区：
文献类型：
--
作者：
Wen-Long Chin
正交频分复用（OFDM）是一种很有前途的通信技术。然而，在分散衰落信道上OFDM的同步仍然是一个重要而具有挑战性的问题。本文提出了一种基于最大似然准则确定OFDM系统中符号时序偏移和载波频偏的同步算法。新的ML方法考虑了时间色散衰落信道，并利用了循环前缀在每个采样时间的独特相关特性。发现所提出的符号时序估计是符号时序偏移和信道长度的二维函数。与以前的ML方法相比，所提出的似然函数在每次采样时都进行了优化，而不需要额外的导频符号。为了实际实现所提出的方法，采用了一种次优的机器学习估计方法，并给出了一个近似但封闭的解。近似解的非线性运算可以使用传统的查找表来实现，以降低计算复杂度。所提出的CFO估计也依赖于信道长度。与传统的方法不同，该方法充分利用了色散衰落信道的延迟扩展（这通常会降低估计的精度）。在此基础上，分析了CFO估计的cram<s:2> - rao下界（CRLB），并通过仿真验证了该估计方法的优越性。
Orthogonal frequency-division multiplexing (OFDM) is a promising technology for communication systems. However, the synchronization of OFDM over dispersive fading channels remains an important and challenging issue. In this paper, a synchronization algorithm for determining the symbol timing offset and the carrier frequency offset (CFO) in OFDM systems, based on the maximum-likelihood (ML) criterion, is described. The new ML approach considers time-dispersive fading channels and employs distinctive correlation characteristics of the cyclic prefix at each sampling time. The proposed symbol timing estimation is found to be a 2-D function of the symbol timing offset and channel length. When compared with previous ML approaches, the proposed likelihood function is optimized at each sampling time without requiring additional pilot symbols. To practically realize the proposed method, a suboptimum approach to the ML estimation is adopted, and an approximate but closed-form solution is presented. Nonlinear operations of the approximate solution can be implemented using a conventional lookup table to reduce the computational complexity. The proposed CFO estimation is also found to depend on the channel length. Unlike conventional schemes, the proposed method fully utilizes the delay spread of dispersive fading channels (which usually reduces the accuracy of estimations). Furthermore, the Cramér-Rao lower bound (CRLB) on the CFO estimate is analyzed, and simulations confirm the advantages of the proposed estimator.