基于最小熵准则的频偏估计算法研究

提出了一种基于SEM(Spectrum Entropy Minimization,频谱熵最小化)的频偏估计算法。研究发现,频偏越小,离散信号频谱的熵越小。根据这一性质建立了相关算法,通过一维搜索的方式得到使相位补偿后信号频谱熵最小的频偏估计。仿真结果表明:SEM算法的估计性能能够达到CR(Cramer-Rao,克拉美罗)下界,而且估计的稳定性和鲁棒性较好,具有一定的应用价值。

A frequency bias estimation algorithm based on spectrum entropy minimization

This paper proposes a frequency bias estimate method based on SEM (Spectrum Entropy Minimization).We found that smaller the frequency bias is,smaller the signal spectrum entropy is.In the algorithm,a phase filter is established to compensate for the frequency bias until the global minimum entropy is achieved and the estimate of the frequency bias is done.Simulation results show that the proposed algorithm reaches CR (Cramer-Rao) lower bound and performs well in robustness.