NOISE CANCELLATION AND SOURCE LOCALIZATION

A noise-canceling processor is developed for matched-field processing problems involving a signal buried in noise. This processor is based on modeling both signal and noise and searching the space of unknown parameters to achieve the best agreement between covariances. The noise-canceling processor reduces to the Bartlett processor in the limit of high signal-to-noise ratio. The examples illustrate the localization of a source obscured by interference from ambient noise or a second source. The noise-canceling processor is also applied to localize a silent object using scattered ambient noise.