基于自适应门限局部改进及求根MUSIC算法

MUSIC-algorithm based on adaptive threshold of part amelioration and extract

使用合适的门限才能正确地提取出小特征值,如果能适应较大范围的信噪比和快拍数就需要有一定自适应能力的门限。通常,信噪比对门限的影响比快拍数对门限的影响要大得多,经过反复计算和仿真得到一个门限公式:(15(SNR)^0.8*(snap)^0.2),其中SNR为信噪比,snap为快拍数。多次仿真表明,该门限公式具有较好的自适应能力,即能很好地适应各种信噪比与快拍数的变化,能较准确地估计出信源个数。

Using the proper threshold will extract the minimum eigenvalue correctly. Besides, it must have self- motion to the gate limit when it could adapt to signal -to -noise ratio of large area and fast snap rate. Usually, signal -to -noise ratio has more influence on the gate limit than fast snap rate does. There exists a gate -limit formula after a lot of experiments, and it is （15 （SNR）0.8 ＊ （snap） 0.2）. The SNR means signal - to - noise ratio and snap means the rate of taking a photograph. After taking repetitious experiments, it shows this gate - limit formula has good adaptive ability, which could adapt to various SNR and change of fast snap rate, also exactly estimate the signal source.