Inter-array coherence-based identification of static acoustic clutter sources

Abstract

In multi-array acoustic detection and tracking applications, there often exist valid acoustic sources that are unwanted because of their position or nature, e.g., generators and air-handling equipment. During a tracking scenario where desired targets can be moving and/or piecewise static, the presence of these clutter sources can negatively impact the localization of a desired target by interrupting a track or generally increasing the noise floor local to an array. In directions away from the clutter source, beamforming methods onboard an array, e.g., minimum variance distortionless response, can be used to suppress the signature of the unwanted source. However, when a desired target passes through the azimuth of a clutter source, beamformers may fail to separate the desired source from the clutter. Here, inter-array coherence between beamformed solutions will be used to identify a static source by its properties in order to remove the clutter source as a potential target in a tracking application. Acoustic data on clutter sources measured by multiple microphone arrays separated by tens of meters will be discussed. In multi-array acoustic detection and tracking applications, there often exist valid acoustic sources that are unwanted because of their position or nature, e.g., generators and air-handling equipment. During a tracking scenario where desired targets can be moving and/or piecewise static, the presence of these clutter sources can negatively impact the localization of a desired target by interrupting a track or generally increasing the noise floor local to an array. In directions away from the clutter source, beamforming methods onboard an array, e.g., minimum variance distortionless response, can be used to suppress the signature of the unwanted source. However, when a desired target passes through the azimuth of a clutter source, beamformers may fail to separate the desired source from the clutter. Here, inter-array coherence between beamformed solutions will be used to identify a static source by its properties in order to remove the clutter source as a potential target in a tracking application. Acous...