Title :
Fast RLS with parallel computations
Author :
Djigan, Victor I.
Author_Institution :
R&D Center of Microelectronics, ELVEES, Moscow, Russia
Abstract :
The paper presents parallel forms of the sliding window fast RLS algorithms with the regularization of adaptive filter correlation matrix. The new algorithms are mathematically identical to the appropriate algorithms with sequential computations, but can be implemented by means of parallel processors. The algorithms are based on the using of the generalized matrix inversion lemma. They can be used for the processing of non-stationary signals.
Keywords :
adaptive filters; adaptive signal processing; filtering theory; least squares approximations; parallel processing; recursive filters; adaptive filter correlation matrix; generalized matrix inversion lemma; nonstationary signal; parallel computation; parallel processor; sliding window fast RLS algorithm; Adaptive filters; Adaptive signal processing; Concurrent computing; Filtering algorithms; Least squares approximation; Linear matrix inequalities; Neural networks; Resonance light scattering; Signal processing; Signal processing algorithms; Adaptive filtering; RLS; fast RLS; parallel computations;
Conference_Titel :
Emerging Technologies: Circuits and Systems for 4G Mobile Wireless Communications, 2005. ETW '05. 2005 IEEE 7th CAS Symposium on
Print_ISBN :
5-7422-0895-2
DOI :
10.1109/EMRTW.2005.195676
