Random sampling applied to array beamforming: The aggregate beamformer
David I. Havelock · The Journal of the Acoustical Society of America · 2002
It is shown how beamforming can be accomplished by randomly sampling array elements and time-aligning the data samples. The time-aligned samples are referred to as the ‘‘aggregate’’ signal. Time-alignment is done using conventional beamforming delays but data values are not summed. Beamforming weights (array shading) can be applied by modifying the sampling probability for each sensor. Simple methods for reconstruction of the original signal from the aggregate signal are discussed. The reconstructed signal of this aggregate beamformer has the same directional response as a conventional beamformer except that off-beam signals are transformed into white noise. This noise can be reduced as required by adjusting the sampling rate of the aggregate signal. Higher beam-steering resolution than that of conventional beamforming is achieved without the need for data interpolation. The hardware and computational complexity of the aggregate beamformer does not increase as the number of array elements is increased. Aggregate beamforming offers a high degree of integration for arrays with many elements, such as 3-D arrays.