Title

Two-Dimensional Frequency Domain, Adaptive System Modeling Using 3-D Spatiotemporal Inputs

Abstract

In this paper, an adaptive, two-dimensional (2-D) frequency domain system modeling technique is presented. This technique is implemented in the 3-D spatiotemporal domain. The algorithm derived here is based on the recursive like structure, and models the unknown 2-D spatially linear and invariant system by a 2-D auto-regressive, moving-average (ARMA) structure. For each pair of input and output images of the 2-D unknown system, the algorithm, which uses the steepest descent method, minimizes the error function in the least-square sense by adjusting the coefficients of the 2-D ARMA model at each iteration. Also, the homogeneous convergence factors associated with this algorithm are developed. Computer simulations demonstrate that this algorithm, with the derived homogeneous convergence factor, has excellent adaptation accuracy and convergence speed. For illustration, this algorithm is successfully applied to modeling a time varying 2-D system.

Publication Date

1-1-1993

Publication Title

Proceedings - IEEE International Symposium on Circuits and Systems

Volume

1

Number of Pages

315-318

Document Type

Article; Proceedings Paper

Identifier

scopus

Personal Identifier

scopus

Socpus ID

0027262080 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/0027262080

This document is currently not available here.

Share

COinS