Paper Info
Title
Performance Analysis of the Normalized LMS Algorithm for Complex-Domain Adaptive Filters in the Presence of Impulse Noise at Filter Input
Abstract
This letter develops theoretical analysis of the normalized LMS algorithm (NLMSA) for use in complex-domain adaptive filters in the presence of impulse noise at filter input. We propose a new "stochastic" model for such impulse noise, and assume that filter reference input process is a white process, e.g., digital QAM data, White & Gaussian process, etc. In the analysis, we derive a simple difference equation for mean square tap weight misalignment (MSTWM). Experiment is carried out to demonstrate effectiveness of the NLMSA in robust filtering in the presence of the impulse noise at the filter input. Good agreement between simulated and theoretically calculated filter convergence, in a transient phase as well as in a steady-state, proves the validity of the analysis.
Year
DOI
Venue
2006
10.1093/ietfec/e89-a.9.2422
IEICE Transactions
Keywords
Field
DocType
digital qam data,filter reference input process,complex-domain adaptive filter,filter input,good agreement,white process,gaussian process,normalized lms algorithm,complex-domain adaptive filters,impulse noise,theoretically calculated filter convergence,theoretical analysis,performance analysis,adaptive filter,lms algorithm
Raised-cosine filter,Digital filter,Linear filter,Control theory,Computer science,Infinite impulse response,Adaptive filter,Finite impulse response,Nonlinear filter,Filter design
Journal
Volume
Issue
ISSN
E89-A
9
0916-8508
Citations 
PageRank 
References 
5
0.47
0
Authors
1
Name
Order
Citations
PageRank
Shin'ichi Koike1174.93