Direction of arrival estimation by subspace methods

It is demonstrated that a Bayesian analysis can be simplified and can lead to a practical direction-of-arrival (DOA) estimator. This estimator is remarkably similar to the MUSIC algorithm, but has a threshold which is significantly lower than that for MUSIC. The reason for this improvement in performance is identified. This result is important for three reasons. First, the method is far superior to MUSIC, and yet is clearly of the same class of methods. Second, the derivation is completely different for the usual subspace arguments used to justify MUSIC. Third, like MUSIC, the method is applicable to arbitrary array geometry