一种子带分频段信道估计算法

针对当前高速率通信中信道阶数很长导致信道估计和均衡困难的问题，利用子带滤波器组近似完全重构的特点，提出一种在子带内进行分频段信道估计、在全频带综合信道参数的估计方法。该方法较全频带信道估计收敛速度快，收敛误差小，能很好适应恶劣的信道情况。虽然总的计算量大于全频带信道估计，但由于采用并行计算，所以能大大减少运算时间。仿真试验表明，在重构误差足够小的情况下，子带数目越多，收敛越快，收敛残差比全频带信道估计小5 dB左右。

A channel estimation algorithm based on subband frequency-domain decomposition

In high speed communications, the channel estimation and equalization are difficult because of the very long channel length. In allusion to the difficult points,by using the near perfect reconstruction speciality of filter banks, an algorithm with channel estimation in every subband and constructing the parameters in fullband is proposed. Compared with the fullband channel estimation, the convergence speed of the proposed method is faster, the convergence error is smaller, and the adaptability for bad channel is better. The whole computation load is more than that of fullband channel estimation, but the computation time is reduced largely because of the parallel computation of subband signals. Computer simulations illustrate the proposed method has faster convergence speed with more subbands and 5 dB convergence error less than the fullband channel estimation, if the reconstruction error is small enough.