基于PN序列的OFDM时频同步改进算法

在训练符号辅助的正交频分复用同步算法中，训练符号中循环前缀的起始位置会出现与正确定时位置处强度相近的相关峰，从而造成定时错误。针对该问题，将现有共轭对称结构的训练符号，乘以伪随机加权因子而构造出新的同步训练符号，基于新的训练符号给出定时同步测度函数，并提出相应的定时同步算法和频偏估计算法。利用伪随机加权因子的自相关性，可有效消除循环前缀的相关峰，减少定时错误，提高频偏估计精度。理论分析和仿真结果表明，该算法在改善同步性能的同时，可降低计算复杂度。

Improved OFDM Timing and Frequency Synchronization Algorithm Based on PN Sequence

The correlation peak located in the initial position of Cyclic Prefix(CP) in training symbol has the similar intensity with the one at the correct timing position, which causes the timing error in Orthogonal Frequency-Division Multiplexing(OFDM) preamble-aided synchronization algorithms. To overcome the problem, a new synchronization training symbol is formed with the conjugated symmetrical training symbol by multiplying Pseudo-Noise(PN) weighted factor, this paper proposes a new timing metric as well as the corresponding algorithm of improved timing synchronization and frequency offset estimation based on the new structure of training symbol. The proposed algorithm eliminates cyclic prefix correlation peak effectively so as to overcome the inaccurate timing and improve the accuracy of frequency offset estimation by utilizing autocorrelation of the pseudo-random weighted factor. Simulation results show that the proposed algorithm not only improves the synchronization performance, but also reduces the computational complexity.