Title :
A new QR-based block least mean squares (QR-BLMS) algorithm for adaptive parameter estimation
Abstract :
This paper proposes a new family of QR-based block LMS (QR-BLMS) algorithms for adaptive parameter estimation. It extends the QR-based LMS (QR-LMS) algorithm by handling a block of data vectors at a time instead of one single input vector. Moreover, it is shown that when there is only one new input vector in the block and the others are obtained from previous time instants, this QR-BLMS algorithm yields a new QR-based implementation of the well-known affine projection algorithm (APA). Simulation results for an acoustic echo canceller showed that the QR-BLMS and QR-APA algorithms perform better than the QR-LMS algorithm.
Keywords :
acoustic signal processing; adaptive estimation; adaptive filters; adaptive signal processing; echo suppression; least mean squares methods; parameter estimation; LMS; QR-BLMS algorithm; QR-based block least mean squares; acoustic echo canceller; adaptive parameter estimation; affine projection algorithm; performance; Arithmetic; Convergence of numerical methods; Echo cancellers; Equations; Least squares approximation; Matrix decomposition; Parameter estimation; Projection algorithms; Transversal filters; Vectors;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03). 2003 IEEE International Conference on
Print_ISBN :
0-7803-7663-3
DOI :
10.1109/ICASSP.2003.1201766
