Variable tap-length algorithm using lattice structured LMS adaptive filters

In a classic implementation of lattice structure adaptive filters, the number of taps are often fixed making it hard to achieve the best potential performance. In this paper, we propose a novel dynamic method to search for the optimum tap-length while adapting the lattice joint process estimator coe...

Full description

Saved in:
Bibliographic Details
Published in2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering (CCECE) pp. 1 - 5
Main Authors Alsaid, Salem, Abdel-Raheem, Esam, Khalid, Mohammed, Mayyas, Khaled
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.04.2017
Subjects
Computers
Convergence
Filtering algorithms
Finite impulse response filters
Lattices
Prediction algorithms
Signal to noise ratio
Online AccessGet full text
DOI10.1109/CCECE.2017.7946702

Cover

More Information
Summary:In a classic implementation of lattice structure adaptive filters, the number of taps are often fixed making it hard to achieve the best potential performance. In this paper, we propose a novel dynamic method to search for the optimum tap-length while adapting the lattice joint process estimator coefficients. The concept of pseudo fractional tap-length (FT) is incorporated in the lattice joint process estimation structure, where the proposed algorithm needs little extra computational complexity in order to attain optimization for both of the adaptive filter coefficients and the filter tap-length.
DOI:10.1109/CCECE.2017.7946702