Updating MUSIC and root-MUSIC with the rank-revealing URV decomposition

The authors describe direction-of-arrival (DOA) algorithms that possess the virtues of subspace-based DOA algorithms but are not expensive to compute or difficult to update. The new DOA algorithms are based on a new decomposition, URV decomposition (URVD), which can be updated in O(N/sup 2/) and serves as an intermediary between QR decomposition and singular value decomposition. The URVD is outlined for updating subspace-based DOA algorithms, specifically MUSIC. It is shown, how to update the URVD and how it may be used for MUSIC. In addition, the authors describe how to update its variant, root-MUSIC. Root-MUSIC is used as a means for statistical performance evaluation. Several simulation results are presented in terms of mean square error and bias of the direction estimates.<
>