资 源 简 介
用Matlab程序实现P阶Levinson-Durbin算法。以一个2阶自回归模型（参数为b0=1, a1=0, a2=0.81）和一个2阶滑动平均模型（参数为b0=1, b1=1, b2=1）为例，选取观测数据长度为1000，分别用一个AR(2)模型和一个AR(10)阶模型来估计其功率谱。设激励信号模型的高斯白噪声的均值为0，方差为1。用Levinson-Durbin算法迭代计算AR模型参数，并用估计出的AR模型参数画出观测信号的功率谱。并对Levinson-Durbin算法的性能进行分析。-Write a small MATLAB program that implements the pthorder Levinson-Durbin (L-D). Run/Test the program using a AR(2) process (b0=1,a1=0, a2=0.81) and an MA(2) (bn=1,1,1) process-about 1000 samples. Use L-D with p=2 (for the AR) and 10 (for the MA). Plot the AR spectra produced in the two cases with L-D. List the direct form and the reflection coefficients in a table. Profile the L-D (total number of computations for a pthorder
