Effect of Sparse Array Geometry on Estimation of Co-array Signal Subspace

摘要

This paper considers the effect of sparse arraygeometry on the co-array signal subspace estimation errorfor Direction-of-Arrival (DOA) estimation. The second largestsingular value of the signal covariance matrix plays an importantrole in controlling the distance between the true subspace andits estimate. For a special case of two closely-spaced sourcesimpinging on the array, we explicitly compute the second largestsingular value of the signal covariance matrix and show that itcan be significantly larger for a nested array when comparedagainst a uniform linear array with same number of sensors.