Title :
A new algorithm for adaptive line spectral pair filter and its convergence behaviour
Author :
Taihui, Jiang ; Youwei, Zhang ; Junying, Gan
Author_Institution :
Inst. of Inf. Sci., Wuyi Univ., Jiangmen City, China
Abstract :
The definition of line spectral pair (LSP) coefficients are first derived, then a least-mean-square (LMS) type algorithm for end-point error is is used to calculate LSP coefficients. Experimental results show that the algorithm has a higher convergence rate and lower misadjustment as compared with the LMS algorithm for an adaptive transversal filter. The LSP coefficients have been proposed as an alternative representation of linear predictive coding (LPC) speech parameter for subband coding techniques with dynamic bit allocation in each subband
Keywords :
adaptive filters; convergence of numerical methods; filtering theory; least mean squares methods; linear predictive coding; spectral analysis; speech coding; LSP coefficients; adaptive line spectral pair filter; adaptive transversal filter; convergence behaviour; convergence rate; dynamic bit allocation; end-point error; experimental results; least mean square algorithm; linear predictive coding; misadjustment; speech parameter; subband coding; Adaptive filters; Convergence; Digital filters; Lattices; Least squares approximation; Nonlinear filters; Predictive models; Signal processing algorithms; Speech analysis; Transversal filters;
Conference_Titel :
Communication Technology Proceedings, 1996. ICCT'96., 1996 International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-2916-3
DOI :
10.1109/ICCT.1996.545216
