基于置信度的抑制信道估计中噪声的方法

Noise suppression in channel estimation based on level of confidence

无线信道中的噪声会造成信道冲激响应的估计误差,降低系统性能。为了有效降低噪声对信道估计结果的影响,提出了一种基于信道冲激响应估计中噪声分布特性的置信度方案。该方案根据信道估计结果的多帧统计特性计算每个采样点的置信度,并根据置信度对信道估计结果中的多径与噪声进行区分,然后对噪声进行有效地抑制。分析和仿真结果表明,该方法能够降低信道估计的误差。在双PN序列填充的时域同步正交频分复用(TDS-OFDM)系统中,其误符号率较传统的信道估计算法在静态和时变信道下均有一定的性能改善,同时该算法也大大降低了硬件实现的复杂度。

The noise in wireless channel generally causes errors of the estimation on the channel impulse response（CIR）,therefore,degrades system performance.A noise suppression method was developed based on the level of confidence according to the noise distribution properties to suppress the noise impact in the estimated CIRs.The algorithm calculates the level of confidence for every sample based on the multi-frame statistical properties of the estimated CIRs,distinguishes the paths and noise in the estimated CIRs with the level of confidence,and then effectively suppresses the noise.Analyses and simulations show that the method can improve the channel estimation accuracy.The method improves the signal-noise ratio（SNR） compared to traditional channel estimation algorithms over static and time-varying channels in the time-domain synchronous orthogonal frequency division multiplexing（TDS-OFDM） system with dual PN-sequence padded.The algorithm also dramatically reduces the hardware complexity.