最大熵线性预测超分辨算法的误差灵敏度分析

Sensitivity Analysis of the Superresolution Algorithm Based on the Maximum Entropy Linear Prediction

本文分析了Burg提出的最大熵线性预测超分辨算法对阵列幅度和相位误差的灵敏度。给出了具有两个目标、不同阵元数和不同间距等情况下的数值计算结果。结果表明最大熵算法较最大似然估计(MLM)算法和MUSIC算法对幅相误差的影响更加敏感。文中还给出了最大熵算法和MUSIC算法分辨目标的实验结果。

This paper analyses the sensitivity of the direction-of-arrival (DOA) to the gain and phase errors, concerned in the superresolution algorithm based on the maximum entropy linear prediction introduced by J. P. Burg, and gives the numerical results related with different number of elements and different element spacing about two sources. The results show that the maximum entropy method is more sensitive to the gain and phase errors than the maximum likelihood method (MLM) and MUSIC method. The experimental results of DOA with maximum entropy and MUSIC algorithm are also presented in the paper.