"JOURNAL OF RADIOELECTRONICS" N 10, 2002 |
A CNLMS ADAPTIVE FILTER ALGORITHM
Casco Fausto, Amín Omar, Zaharov Viktor and Ekzhanov Alexander.
Tel (5255) 58044637 Fax (5255) 5804 4628 Email:alfa@xanum.uam.mx
key words : System Identification and Parametric Spectrum Estimation
Department of Electrical Engineering,
Universidad Autónoma Metropolitana-Iztapalapa,
México City, México.
Received 15.10.2002.
This paper propose a Cascaded structure for adaptive digital filtering used with the Normalized Least Mean Square Algorithm (CNLMS) that can be obtained by the expansion of Laguerre´s polynomials, this structure is implemented as a system identifier and is compared with filter structure using the Normalized Least Mean Square Algorithm (LMS) [6] and with the Variable Step Size Algorithm (MVSS) described in [7]. The number of analytical operations is below of the required for the MVSS algorithm. Simulations show that the new architecture decrease the mean square error (MSE) compared with results obtained of comparison with both NLMS and MVSS algorithms, see figures 3 and 4.