针对基于扩展模型(BEM)的正交频分复用(OFDM)系统时变信道估计中频谱利用率不高的问题,提出了一种多符号BEM方法。相对于传统的单符号BEM方法,提出的方法减少了导频的子载波,并且通过基线倾斜技术抑制了傅里叶级数展开时出现的Gibbs现象。理论分析表明,基线倾斜技术引入后的信道模型误差在归一化多普勒频偏不大于((1/2)-(1/M))/(1+(N_g/N))时均有减小。数值仿真显示在相应的时变信道条件下,提出的方法的估计误差性能和误比特率(BER)性能相对传统方法有明显改善。
In view of the problem of low frequency spectrum utilization for the time-varying channel estimation in orthogonal frequency division multiplexing(OFDM)systems based on basis expansion model(BEM),a new method based on multiple symbols BEM was proposed.Compared with the conventional method based on single symbol,the proposed method reduced the subcarriers for pilots.And by means of the baseline tilting technique,the new method mitigated the Gibbs phenomenon caused by Fourier series expansion.The theoretical analysis reveals that the channel model error is reduced due to the introduction of the baseline tilting technique when the normalized Doppler frequency is no more than ((1/2)-(1/M))/(1+(N_g/N)).Finally,the numerical simulation shows that compared with conventional methods,the proposed algorithm achieves significant improvement on the estimation error performance and bit error rate(BER)performance under the conditions of the corresponding time-varying channel.