A General Approach for Operational Bandwidth Extension in Spherical Microphone Array

Abstract

Background and Objectives: Operating frequency range of a microphone array is limited by the array configuration. Spatial aliasing occurs at frequencies considered to be out of the microphone array operating range that leads to side-lobes in the array beam pattern and consequently degrades the performance of the microphone array. In this paper, a general approach for increasing the operational bandwidth of the spherical microphone array without physical changes to the microphone array is proposed. Methods: Recently, Alon and Rafaely proposed a beamforming method with aliasing cancellation and formulated it for some well-known beamformers such as maximum directivity (MD), maximum white noise gain (WNG), and minimum variance distortionless response (MVDR) which have been called MDAC, MGAC, MVDR-AC beamformer respectively. In this paper, we derive MDAC method from different point of view. Then, based on our perspective, we propose a new method that is easily applicable for any beamforming algorithms.Results: Comparing with MDAC and MGAC beamformers, performance measures for our approach show improvement in directivity index (DI) and white noise gain (WNG) by nearly 19% and 15% respectively.Conclusion: Aliasing and, in consequence, unwanted side lobe formation is the main factor in spherical microphone arrays operational bandwidth determination. Most of the methods previously presented to reduce aliasing demanded physical changes in the array structure which comes at a cost. In this paper we propose a new method based on Alon and Rafaely’s approach via designing a constrained optimization problem using orthogonality property of spherical harmonics, to achieve better performance.

Keywords

• spherical microphone arrays
• aliasing
• beamforming
• operational bandwidth