Fast updating of maximum likelihood direction of arrival estimates

The author demonstrates that asymptotically efficient direction of arrival estimates can be obtained without resorting to an eigendecomposition. In particular, an algorithm similar in form to weighted subspace fitting and possessing identical asymptotic properties is developed that requires computation of only the sample covariance and its inverse. The algorithm achieves computationally efficient estimate updates, and may be easily parallelized. In addition to describing the algorithm and its asymptotic properties, a simulation example to validate the algorithm´s performance is included