Gradient based processing for linear vector sensor

The demand for higher performance arrays, coupled with recent advancements in vector sensor technology, has resulted in a surge in research on vector sensor based arrays in the past decade. In this paper, a new approach to vector sensor array processing is proposed. The method relies on extracting the so called 'acoustic modes' of the propagating wavefield using the spatial gradients of the field variables. These gradients are approximated with a finite difference scheme along a uniform linear array of vector sensors. The proposed method is evaluated through a series of numerical simulations. The resulting broadband beamformer is shown to have a directivity factor comparable to conventional beamformers with much larger apertures while providing all the advantages of acoustic vector sensors such as port-starboard discrimination. Owing to the sufficiency of shorter apertures and other potential operational advantages, the proposed method is suitable for use in short towed arrays with autonomous underwater platforms.