OFDM系统中改进的FFT插值算法

An Improved FFT channel interpolation in OFDM systems

摘要:

提出了一种基于快速傅立叶变换的边缘扩展信道插值算法（EFFT，Edged Extend Fast Fourier Transform）。该算法对OFDM符号有效频带边缘的两端频点做镜像线性扩展，使增加的导频信息能改善有效频带边缘附近的信道估计性能。仿真结果表明，在低阶调制且SNR较低时，EFFT插值算法的性能与维纳滤波信道估计算法相当，且较传统FFT插值算法的MSE提升1dB左右。在实现复杂度方面，EFFT算法与FFT算法相当；与维纳滤波算法相比运算量明显降低。EFFT算法不需要对现有技术标准进行改动，也不会降低系统的传输效率。

Abstract:

An Edged Extend Fast Fourier Transform (EFFT) channel interpolation algorithm is proposed. The presented EFFT algorithm can extend linear mirror pilots nearby the edge of the effective frequency band of each OFDM (Orthogonal Frequency Divided Multiplexing) symbol and the performance of channel estimation is enhanced by the added mirror pilots. The simulation results show that the performance of the EFFT channel interpolation algorithm is similar to that of Wiener filter channel estimation algorithm under the condition of low SNR and modulation mode, while there was about 1dB improvement when compared to traditional FFT channel interpolation algorithm. The complexity of the EFFT channel interpolation was similar to that of traditional FFT channel interpolation algorithm and there is a significant reduction of calculations when compared to Wiener filter channel estimation algorithm. Moreover, the EFFT channel interpolation is suitable for the existing standards and has no influence on the transmission efficiencies.