一种基于相关序列的OFDM频偏估计算法

提出了一种改进的整数倍频偏估计算法,此算法基于已知频域前导序列及其共轭移位序列的Hadamard积构建一个具有良好自相关性能的相关序列,再利用相关运算估计频偏.利用蒙特卡洛仿真方法,分析了新算法的测度函数在高斯白噪声信道条件下的均值和方差分布特性.将新算法应用于IEEE 802.16d系统中进行仿真,结果表明,该算法能够克服大定时偏移带来的影响,在信噪比大于10 dB时具有精确的整数倍频偏检测概率,且算法复杂度没有明显提高.

Frequency Offset Estimation for OFDM by Using a Correlation Sequence

Based on the Hadamard product of known frequency domain preamble sequence and its conjugate shift sequence, a new sequence with much better auto-correlation property was constructed. Combined with the arithmetic of cross-correlation, an improved integer frequency offset (IFO) estimation algorithm was proposed. Through Monte-Carlo simulations, the distribution of the expectations and variances of the metric function of newly developed algorithm was analyzed over AWGN channels. The proposed algorithm has been applied in IEEE 802.16d system and simulations were carried out. Simulation results show that the newly proposed method can overcome the serious disturbance introduced by large timing offset. When the SNR is greater than 10 dB, the detection probability of IFO can be accurately obtained without increasing the complexity of calculation.