Broadband beamspace DOA estimation algorithms

Two kinds of broadband beamspace high-resolution DOA estimation algorithms are studied in this paper. The first one is broadband beamspace coherent high-resolution processing method based on constant beamwidth beamformers. In this method, conventional narrowband beamspace high-resolution DOA estimation methods such as beamspace MUSIC can be applied to the output of constant beamwidth beamformers. Because of the fixed beamformers used in this method, its performance will be reduced when there exist strong interference signals outside the beam coverage region. To solve this problem, broadband incoherent high-resolution processing method based on subband beamformers is proposed in this paper. The broadband signals are divided into subbands via frequency decomposition. In each subband, we use MVDR beamformers to suppress the strong interference signals outside the beam coverage region. To overcome the beampattern distortion problem due to the inaccuracy estimation of the covariance matrix, diagonal loading is applied to the MVDR beamformers. Beamspace MUSIC is used to get the spatial spectrum of each subband. The final DOA estimates result from the incoherent combination of spatial spectrum of each subband. Simulation results are presented to show the performance of these two algorithms.