Abstract: In the light of new challenges facing frequency synchronization in Non-continuous Orthogonal Frequency Division Multiplexing (NC-OFDM) systems, a Carrier Frequency Offset(CFO) estimation scheme based on Zadoff-Chu transform matrix is proposed. The transmitted data enjoys the constant envelop property with the help of a Zadoff-Chu transform block in the transmitter, thus the estimation of CFO can be obtained through minimizing the power difference on the same subcarriers between two consecutive symbols. After the frequency compensation, original signals can be restored by multiplying with the inversion Zadoff-Chu matrix. The identifiability problem is proved. Numerical simulations show that the proposed estimator provides an improved Bit Error Rate(BER) compared with existing blind methods.

Key words: cognitive radio, Non-continuous Orthogonal Frequency Division Multiplexing(NC-OFDM), frequency offset estimation, carrier frequency offset, Zadoff-Chu sequence, identifiability problem

摘要： 针对非连续正交频分复用(NC-OFDM)系统中频率同步的干扰问题，提出一种利用Zadoff-Chu变换矩阵的频偏估计算法。发送机的Zadoff-Chu矩阵变换模块将信号转换为恒定包络，接收机通过最小化2个相邻接收NC-OFDM符号中相同子载波的能量差获得频偏估 计，将频率补偿接收信号与Zadoff-Chu逆矩阵相乘，恢复出原始发送信号。仿真结果表明，该算法比其他算法有更好的误码率性能。

关键词: 认知无线电, 非连续正交频分复用, 频偏估计, 载波频偏, Zadoff-Chu序列, 可分辨性问题

CLC Number: 

• TP393