POLYNOMIAL BASED RECURSIVE AND NON RECURSIVE FILTER DESIGN

Abstract

In this dissertation, several algorithms to design linear phase Finite Impulse Response (FIR) and Infinite Impulse Response (IIR) filters have been discussed. Contrary to various already existing standard algorithms,the proposed methods approximate magnitude and phase characteristics simultaneously. The basic mechanism used in this study is polynomial based design of digital filters. We have used several already existing polynomials; e.g., Chebyshev polynomials, Legendre polynomials, to develop linear phase digital filters and developed some two dimensional polynomials following orthogonal properties to design digital filters for image processing, their design methodology have also been discussed. Filters of proposed type can be used for applications where exact linear phase is required. Another application of this type of filters is the design of filters with zero group delay. IIR filters are designed with absolute linear phase and zero group delay. The algorithms proposed in the present thesis allow user to design filters with his set of constraints, which is required in practical filter design problems. Very narrow band 1D and 2D linear phase FIR filters can eas- ily be designed by the proposed methodology. The IIR filters proposed provide the guarantee to result in a stable filter. All the algorithms have been discussed stepwise to make sure that any one with basic programming capability can easily design them. We have not used any standard routine of any particular platform, therefore any freely available programming platform (like C, C++, Scilab, Octave, etc.) can be used to design these filters.