We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Improved clipped affine projection adaptive algorithm.
- Authors
Bekrani, Mehdi; Bibak, Ruhollah; Lotfizad, Mojtaba
- Abstract
In this study, a modification to the affine projection adaptive filtering algorithm is proposed which is based on a three-level clipping of the input signal as applied to the weight update process. This clipping operation causes the input signal to be quantised into three levels, namely, 0, 1, and -1. By doing so, the proposed scheme, in addition to reducing the computational complexity, achieves an error convergence performance that is comparable to, or even at times better than that of the conventional affine projection algorithm (APA) for a certain range of the clipping threshold. The proposed adaptive algorithm is compared with some low complexity variants of APA with respect to the convergence rate and computational complexity, and its superiority to its counterparts of the same order of complexity in terms of the convergence rate is demonstrated. The mean square error analysis based on the energy conservation relation and the stability analysis is also presented for the proposed clipped APA.
- Subjects
ADAPTIVE filters; SIGNAL processing; COMPUTATIONAL complexity; ERROR analysis in mathematics; ALGORITHMS
- Publication
IET Signal Processing (Wiley-Blackwell), 2019, Vol 13, Issue 1, p103
- ISSN
1751-9675
- Publication type
Article
- DOI
10.1049/iet-spr.2018.5216