一种有效的短序列功率谱估计算法及其应用

An Efficient Short Sequence Power Spectrum Estimation Method and Its Application

由于实际中采集到的数据量总是有限的 ,所以要准确获得待研究数据的特性必须采用高效的短序列估计方法。本文提出了一种基于伯格算法的新的短序列谱估计方法 ,该方法在阶数估计时引入收敛因子 ,从而更为有效地估计了阶数 ;同时不直接用伯格算法计算反射系数 ,而是先求得二阶预测误差系数 ,继而求一阶反射系数 ,再进一步求得高阶系数来实现谱估计。计算机仿真表明 ,这种算法可以有效地减小谱偏 。

In practice,only limited data can be obtained so it is very necessary to use an efficient estimation algorithm to get exact characters of the short data sequence. Here we use an improved Burg algorithm to analyse the experimental data. A convergent factor is used to efficiently improve the order estimation. In this case, the first order reflecting coefficient is not directly obtained by Burg algorithm but by solving the coefficients of second order PEF according to making the total output power be minimum. After obtaining the coefficients of the second order PEF, the first order and the other higher order coefficients of PEF can be calculated. The simulation results indicate that the peak shifting of spectrum is efficiently reduced and the resolution is improved.