Abstract: To frequency and timing offsets problem caused by propagation delay and Doppler effect in OFDM wireless systems，based on correlation of training symbol sequence，an efficient timing offsets estimation algorithm called CBT is proposed.To improve the ability of anti-timing offsets，the proposed CBT constructs a simple training symbol sequence，and estimates timing offsets by correlation of the training symbol sequence to find the starting point of the symbol at the maximum point of the timing metric.Detailed simulation results and comparison with an existed PML algorithm prove that the proposed CBT algorithm obviously decreases the errors of timing offsets in AWGN and Rayleigh-fading channels.

Key words: Orthogonal Frequency Division Multiplexing（OFDM）, timing offset estimation, symbol sequence

摘要： 针对受到传播时延、多普勒效应等因素影响的正交频分复用（OFDM）无线系统时间偏移问题，提出了一种基于训练符号序列相关性的时间偏移估计算法CBT。该算法通过对训练符号序列的简单构造，利用其相关性通过时间尺度的最大值找到符号的起始点，实现时间偏移估计，进而有效地提高OFDM系统抗时间偏移的能力。详细的仿真实验以及与现有PML算法的性能比较表明，提出的时间偏移估计算法CBT在加性高斯白噪声（AWGN）与瑞利衰落信道下都明显地减小了时间偏移误差。

关键词: 正交频分复用, 时间偏移估计, 符号序列