Title :
A finite memory non stationary LMS algorithm for adaptive tracking Markovian time-varying channel
Author :
Turki, M. ; Jaidane-Saidane, M.
Author_Institution :
ENIT, LS Telecoms, Tunis, Tunisia
Abstract :
This paper presents a new adaptive algorithm designed to track stochastic time-variations of channels that are characterized by P-order non stationarity Markov model. The proposed Finite Memory Non Stationary LMS (FM-NSLMS) algorithm, not only is able to identify the unknown order and parameters of the Markov model as the NSLMS but also takes into account the recursive nature induced by the non stationarity. The proposed FM-NSLMS performs better than the NSLMS since it has some memorization capability
Keywords :
Markov processes; adaptive estimation; adaptive filters; filtering theory; least mean squares methods; parameter estimation; time-varying channels; tracking; Markovian time-varying channel; P-order non stationarity Markov model; adaptive algorithm; adaptive tracking; finite memory nonstationary LMS algorithm; memorization capability; stochastic time-variations; Adaptive algorithm; Algorithm design and analysis; Convergence; Filters; Least squares approximation; Recursive estimation; Statistics; Steady-state; Stochastic processes; Time-varying channels;
Conference_Titel :
Electronics, Circuits and Systems, 1998 IEEE International Conference on
Conference_Location :
Lisboa
Print_ISBN :
0-7803-5008-1
DOI :
10.1109/ICECS.1998.813362
