Title :
Solution of block Toeplitz equations by matrix iterative methods
Author_Institution :
Dept. of Electr. & Comput. Eng., US Naval Postgraduate Sch., Monterey, CA, USA
Abstract :
Summary form only given. An algorithm for the solution of block matrix equations has been developed, using some well-known matrix iterative techniques. The algorithm requires fewer computations than the direct matrix inversion methods and is very simple to implement. The problem of solving block matrix equations of the Toeplitz form arises in several areas of signal processing such as array processing, optimal filtering, and spectral estimation of two-dimensional signals with both quarter-plane and asymmetric half-plane supports, and in one- and two-dimensional multichannel spectral estimation. Efficient solution methods for these equations are of considerable interest
Keywords :
filtering and prediction theory; iterative methods; matrix algebra; parameter estimation; picture processing; signal processing; array processing; asymmetric half-plane supports; block Toeplitz equations; block matrix equations; image processing; matrix iterative methods; multichannel spectral estimation; optimal filtering; quarter-plane supports; signal processing; two-dimensional signals; Array signal processing; Autocorrelation; Computer simulation; Convergence; Equations; Filtering; Iterative algorithms; Iterative methods; Signal processing; Signal processing algorithms;
Conference_Titel :
Multidimensional Signal Processing Workshop, 1989., Sixth
Conference_Location :
Pacific Grove, CA
DOI :
10.1109/MDSP.1989.97092