Title :
Degenerate eigenvalues — A method to design adaptive discrete time wavelets
Author :
Stefanoiu, Dan ; Tabus, Ioan
Author_Institution :
Signal Process. Lab., Tampere Univ. of Technol., Tampere, Finland
Abstract :
This paper introduces a new method for designing dyadic filter banks with optimum coding gain. We conjecture that in the generic case, the optimum impulse response is the maximum eigenvector of a certain Toeplitz matrix which is maximally degenerated, which ensures that the maximum eigenvector of the Toeplitz matrix has maximum possible multiplicity. We derive analytically the optimum FIR filter of order N = 6, whereas closed form solutions have been available only up to order N = 4.
Keywords :
FIR filters; channel bank filters; discrete wavelet transforms; eigenvalues and eigenfunctions; FIR filter; Toeplitz matrix; adaptive discrete time wavelet design; coding gain; degenerate eigenvalues; dyadic filter banks design; finite impulse response filter; impulse response; multiplicity; Continuous wavelet transforms; Time-frequency analysis;
Conference_Titel :
Signal Processing Conference, 2000 10th European
Conference_Location :
Tampere
Print_ISBN :
978-952-1504-43-3