Performance study of conditional and unconditional direction-of-arrival estimation

A numerical and analytical study of conditional and unconditional direction-of-arrival (DOA) estimation is presented. Explicit expressions for the unconditional Cramer-Rao bounds on the DOA estimation accuracy and the covariance matrix of the conditional maximum likelihood method are given. It is shown that many DOA estimation methods have the same asymptotic statistical properties under conditional and unconditional models. The situation of two narrowband plane signals impinging on a uniformly spaced linear array is discussed.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>