Title :
An algorithm for polynomial matrix SVD based on generalised Kogbetliantz transformations
Author :
McWhirter, John G.
Author_Institution :
Sch. of Eng., Cardiff Univ., Cardiff, UK
Abstract :
An algorithm is presented for computing the singular value decomposition (SVD) of a polynomial matrix. It takes the form of a sequential best rotation (SBR) algorithm and constitutes a generalisation of the Kogbetliantz technique for computing the SVD of conventional scalar matrices. It avoids “squaring” the matrix to be factorised, uses only unitary and paraunitary operations, and therefore exhibits a high degree of numerical stability.
Keywords :
numerical stability; polynomial matrices; singular value decomposition; Kogbetliantz technique; generalised Kogbetliantz transformations; numerical stability; paraunitary operation; polynomial matrix SVD; scalar matrices; sequential best rotation algorithm; singular value decomposition; unitary operation; Jacobian matrices; MIMO; Matrix decomposition; Polynomials; Signal processing; Signal processing algorithms; Singular value decomposition;
Conference_Titel :
Signal Processing Conference, 2010 18th European
Conference_Location :
Aalborg
