Holland K., Nelson P.

The use of arrays of microphones to locate and characterise acoustic sources is becoming more commonplace. Beamforming algorithms are applied to the microphone outputs to focus the array at a point, or to scan an area or volume for sources. The number and position of the microphones in a beamforming array are determined only by the frequency range of interest and the required spatial resolution. However, these beamforming algorithms cannot yield realistic estimates of source strength when more than one source is present within the ‘beam’ of the array. If realistic estimates of source strength distribution are required, more advanced signal processing algorithms, such as the inverse method, are required. In these cases, it is accepted wisdom that the number of microphones must equal or exceed the number of sources present. For many potential applications of these inverse methods, the size and complexity of the source regions of interest can lead to the requirement for prohibitively large numbers of microphones. It is the purpose of this paper to demonstrate how, through re-formulation of the inverse method, the requirements for the minimum number of microphones can be relaxed when certain assumptions can be made concerning the correlation structure of the source region.