A Fast Irregular Microphone Array Design Method Based on Acoustic Beamforming

There is a recognized need for seeking an optimal microphone array to identify the sound source accurately and promptly. Recent studies have shown that a well-designed array displays superior performance in the source image map. However, these methods are complicated and time-consuming to implement. This article aims to simplify the methods and design a feasible array within the constraints of the microphone number and available area. An irregular array designed by the eigen distance among microphones is proposed to break through the beamforming limitation. Compared with the traditional regular array by both numerical simulations and experiments, it possesses balanced characteristics over a wide frequency range of interest. As expected, all arrays almost point to the sound sources but the unique spacing array improves multisource localization accuracy and achieves small localization deviation. Furthermore, it dramatically reduces the computing time of the deconvolution beamforming algorithm.