Implementation of adaptive array algorithms

From MaRDI portal





The author discusses recursive (on-line, updating) implementations of the minimum variance distortionless response (MVDR) in optimal beamforming and of some high-resolution methods based on the spectral decomposition of the signal covariance matrix. For the MVDR, since the update method based on the Sherman-Morrison-Woodbury formula can be numerically unstable, the author proposes three other new (stable) methods. In updating the spectral decomposition, efficient procedures for updating eigenvalues and singular value decompositions are used. Table II gives complexity and stability results of the analyzed methods for beamforming algorithms: matrix inverse lemma, the proposed matrix inversion, the conjugate direction method, extensions of the Cholesky factor updating. Table III gives complexity results for the fast computational updating methods used in spectral decomposition. Stability is not (yet) analyzed. A special remark: the proposed methods are (or have already been) amenable by systolic arrays, reducing the updating time. The results in the paper are very useful, clearly presented, deserving the attention and interest of the people working in the field.











This page was built for publication: Implementation of adaptive array algorithms

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3773183)