针对低信噪比(≤0dB)下SNR估计的难题,提出了基于低密度奇偶校验码(LDPC)译码辅助的迭代SNR估计算法。该算法先采用期望最大(EM)原理及LDPC译码软信息实现SNR粗估计,再以不同SNR下LDPC软信息硬判结果满足校验矩阵约束程度的差异为判决依据,实现基于判决反馈的SNR精估计。仿真表明,该算法能以相对较小的计算复杂度,使LDPC编码系统在低SNR下获得了较高精度的SNR估计。
It is difficult to estimate signal-to-noise ratio (SNR) at low SNR (~OdB). An iterative SNR estimation algorithm based on LDPC decoding is proposed. By utilizing the expectation-maximization (EM) principle and LDPC decoding, this algorithm could get rough SNR estimation. Then the difference of the code constraint feedback of the hard decision from LDPC decoding with two different SNRs is used as the errors in a decision feedback method, thus obtaining accurate SNR estimation. Simulation results indicate that this algorithm could realize accurate SNR estimation for the LDPC coded systems at the cost of fairly less computation complexity.