Observer-Based Beamforming Algorithm for Real-Time Acoustic Array Signal Processing
Long Bai, Xun Cheng Huang, Edward Peers · 2012
In the field of noise identification with microphone arrays, the conventional delay-and sum (DAS) beamforming is the most popular signal processing technique. However, acoustic imaging results that are generated by conventional beamforming are easily in uenced by background noise, particularly for in situ wind tunnel tests. In addition, conventional beamforming based on arithmetic averaging fails to deliver real-time computational capability. To address the above issues, an observer-based approach is introduced in this paper. This so-called observer-based beamforming method has a recursive form similar to the state observer in classical control theory, thus holds a real-time computational capability. Furthermore, coherent background noise can be removed in iterations. The formulations of observer and the associated algorithm have been discussed in this paper. Two numerical simulations demonstrate the good coherent background noise rejection and real-time computational capability of the observer-based beamforming. The algorithm has also been verified in an experimental case.