Sparse spatial spectral estimation in directional noise environment

White Gaussian noise assumption is widely used in the field of spatial spectral estimation. But, the practical ambient noise usually does not satisfy this assumption, which may seriously affect the performance of direction of arrival estimation. This letter presents a linear noise model that can provide a more complete description of the practical ambient noise field, and then proposes a directional noise field sparse spectrum fitting (DN-SpSF) algorithm to estimate the spatial spectral in a directional noise field. Simulations and experimental results demonstrate the good performance of the DN-SpSF algorithm in a spatially directional noise environment.

from #Audiology via xlomafota13 on Inoreader http://ift.tt/2ceBond
via IFTTT