A multi-layer 2D adaptive filtering architecture based on McClellan transformation

Author

Liu, K.J.R. ; Wu, An-Yeu Andy

Author_Institution

Electr. Eng. Dept., Maryland Univ., College Park, MD, USA

fYear

1993

fDate

3-6 May 1993

Firstpage

1999

Abstract

A fully pipelined systolic array structure for multidimensional adaptive filtering is proposed. It utilizes the McClellan transformation (MT) to reduce the total number of parameters used in the 2D filter. A new multilayer triangular array, which is based on QR-decomposition recursive least squares (RLS) (QRD-RLS) as well as the projection method, are derived for the ´1-D prototype filter´ of MT. The hardware complexity for the new architecture is only O(N). The system latency is also reduced from O(N) to O(log₂N). The QRD-RLS algorithm is suitable for real-time image processing such as video signal processing.

Keywords

adaptive filters; least squares approximations; pipeline processing; recursive filters; systolic arrays; two-dimensional digital filters; video signal processing; McClellan transformation; QR-decomposition recursive least squares; hardware complexity; multi-layer 2D adaptive filtering; multidimensional adaptive filtering; pipelined systolic array structure; projection method; real-time image processing; system latency; triangular array; video signal processing; Adaptive filters; Delay; Hardware; Least squares methods; Multidimensional systems; Nonhomogeneous media; Prototypes; Resonance light scattering; Signal processing algorithms; Systolic arrays;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 1993., ISCAS '93, 1993 IEEE International Symposium on

Conference_Location

Chicago, IL

Print_ISBN

0-7803-1281-3

Type

conf

DOI

10.1109/ISCAS.1993.394145

Filename

394145