宽带DOA估计的类MUSIC波束形成算法

摘要

The MUSIC-like beamforming method (MBM) is presented to improve the wideband direction of arrival (DOA) estimation performance of the frequency domain beanforming method. The wideband signal is first disparted to several narrowband signals in the frequency domain. Then the summation of the MBM spectrum at each narrowband is taken as the estimated spatial spectrum. Consequently, the peak locations provide us the wideband DOA estimation. The MBM deserves a frequency-invariant beampattern in a wide frequency band with equivalent computation load compared with that of the conventional beamforming (CBF) .The simulation results show that the wideband DOA estimation pefformance and resolution ability of the MBM fall in between the incoherent CBF and incoherent MUSIC.%为了提高频域波束形成的宽带波达方向估计性能,提出了类MUSIC波束形成算法(MBM,MUSIC-like Beamfomfing Method).在频域将宽带信号划分为若干窄带信号,叠加各窄带的MBM算法的空间谱后其峰值对应角度即为宽带波达方向估计结果.MBM算法的主瓣宽度在不同分析频率下基本保持不变,计算量与常规波束形成(CBF,Conventional Beamfoming)相当.仿真结果表明,MBM算法的宽带波达方向估计性能和角度分辨能力介于分别叠加各窄带的CBF和MUSIC算法估计结果的ICBF(Incoherent CBF)和IMUSIC(Incoherent MUSIC)算法之间.