For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Convergence Analysis of Adaptive Filters Using Normalized Sign-Sign Algorithm
IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 2005/11/01
Print ISSN: 0916-8508
Type of Manuscript: LETTER
Category: Digital Signal Processing
convergence analysis, adaptive filter, normalized sign-sign algorithm,
Full Text: PDF>>
This letter develops convergence analysis of normalized sign-sign algorithm (NSSA) for FIR-type adaptive filters, based on an assumption that filter tap weights are Gaussian distributed. We derive a set of difference equations for theoretically calculating transient behavior of filter convergence, when the filter input is a White & Gaussian process. For a colored Gaussian input and a large number of tap weights, approximate difference equations are also proposed. Experiment with simulations and theoretical calculations of filter convergence demonstrates good agreement between simulations and theory, proving the validity of the analysis.