GEESE (GEneralized Eigenvalues utilizing Signal subspace Eigenvectors) - A new technique for direction finding

A novel technique for estimating the directions of arrival of multiple signals utilizing the generalized eigenvalues associated with certain matrices generated from the signal subspace eigenvectors is reported. This is based on the well-known property that in the case of an uncorrelated and identical noise field, the subspace spanned by the true direction vectors is identical to the signal subspace. A first-order perturbation analysis is carried out to evaluate the performance of this scheme, when the array output cross-covariances are estimated from the data. It is shown that the angle-of-arrival estimator in its least favorable form is unbiased and has nonzero variance in a two-source scene.