基于噪声功率估计的FMT时域信道估计

利用PN码的循环相关性进行信道估计时,其估计值是有偏的,并且在低信噪比下受信道噪声的影响大。鉴于此,提出了基于信道噪声功率估计的FMT无偏信道估计算法,利用估计值数据特点,完全消除了残余直流偏置,同时在接收端利用导频序列估计信道噪声功率,根据路径信噪比选择出受噪声干扰大而能量贡献小的多径分量并将其置零,降低了信道噪声对估计精度的影响,改善了信道估计性能。

Time-domain FMT Channel Estimation Algorithm Based on Noise Power Estimation

Channel Estimation could be performed based on the characteristic of PN periodic correlation. But the estimation was biased and influenced largely by channel noise when SNR was low. Thereby a new unbiased FMT channel estimation algorithm grounded on channel noise power estimation was proposed. Using the characteristic of estimated parameters, remaining DC bias was eliminated completely. Moreover the channel noise power was estimated utilizing in the pilot sequence. Using the estimated noise power, multipath weights, whose energy was small and which were interfered seriously by channel noise, were selected to be set to zeros. The algorithm reduced the influence of channel noise on estimating precision and improved the performance of channel estimation.