Stable DOA Estimation with Sparse Sensor Arrays

Some recent works on direction of arrival (DOA) estimation using sparse arrays have centered on a difference co-array approach, where DOA estimation is performed with respect to the source variances rather than the source complex amplitudes. This formulation advertises big gains in the available degrees of freedom (DOF) which makes it of interest to sparse arrays. Here we seek to quantitatively compare four super-resolution DOA estimation methods as applied to the difference co-array formulation. Our study assesses performance in terms of stability with respect to the number of data snapshots used and signal-to-noise ratio (SNR). While none of the four methods presented here are able to fully exploit the DOF offered by the difference co-array, we do find that two of the four methods however do exhibit significantly more stable performance in terms of snapshots and SNR.