中文摘要：

在同时同频全双工（CCFD）中,基站发射、接收通道非线性引入的附加干扰以及基站周围环境的变化,会影响数字域自干扰信道估计精度。该文提出了一种变窗长离散傅里叶变换信道估计算法,算法根据实际自干扰信道特征选取最优干扰抑制窗长,提高自干扰信道的估计精度。仿真结果表明：在自干扰信道为莱斯信道,干噪比为15 dB时,采用该算法的数字自干扰抑制能力为22 dB,比采用最小二乘信道估计高7 dB,比采用固定窗长离散傅里叶变换信道估计高3 dB。

英文摘要：

In the co-time co-frequency full duplex （CCFD）, the nonlinear effects of transmit-receive path as well as environment change would change the channel estimation accuracy. A self-interference channel estimation algorithm based on discrete Fourier transform （DFT） with variable window length is introduced. The algorithm can select the best window size according to the channel characteristics and therefore improve the self-interference channel estimation accuracy. Simulation shows that the digital self-interference cancellation with this proposed algorithm is 22 dB when interference-to-noise ratio is 15 dB, which is 7 dB and 3 dB higher than least squares （LS） algorithm and the traditional DFT-based algorithm, respectively.