Comparison of beamforming methods to reconstruct extended, partially-correlated sources

Advanced cross-beamforming methods improve upon traditional beamforming to reconstruct complex source information and to estimate their respective acoustic radiation. Regularization of the cross-beamforming matrix as part of the calculation procedure helps improve method robustness, but differences in implementation impact volume velocity source results and subsequent field predictions. This paper compares the abilities of four regularization-based, cross-beamforming methods: hybrid method, functional beamforming, generalized inverse beamforming, and mapping of acoustic sources (MACS), along with ordinary cross beamforming, in their ability to reproduce source and field characteristics for an extended, partially correlated numerical source that mimics key characteristics of supersonic jet noise radiation. The four methods that rely on regularization significantly outperform cross-beamforming results, yet the effectiveness of each of four methods is dependent on the individual regularization schemes, which...