Eigenfilters: A New Approach to Least-Squares FIR Filter Design and Applications Including Nyquist Filters Advisor : Yung-An Kao Student : Chih-Wei Chen 2006/05/05
IEEE Transaction on circuits and system, vol. CAS-34, NO. 1, January 1987 P.P. VAIDYANATHAN, and TRUONG Q. NGUYEN
Outline Introduction Linear phase FIR Low-Pass eigenfilters Example
Introduction A new method of designing linear-phase FIR filter is proposed, the method is based on the computation of an appropriate real, symmetric, and positive-definite matrix. The proposed design procedure is general enough to incorporate both time and frequency domain constraints Application Nyquist filter Equiripple filter
Introduction The desired response is The amplitude response of H(z) is Type I filter
Introduction The least-squares (LS) approach Linear equation, LSE solution can be express matrix from
Linear phase FIR Low-Pass eigenfilters We wish minimizing an error measure using another method If error measure can be expressed the from
Linear phase FIR Low-Pass eigenfilters The FIR linear phase filter frequency response Type I filter Type II filter
Linear phase FIR Low-Pass eigenfilters Matrix from
Linear phase FIR Low-Pass eigenfilters Stopband error
Linear phase FIR Low-Pass eigenfilters Passband error It cannot be written in the form Change, derive zero-frequency response is given by
Linear phase FIR Low-Pass eigenfilters
Total measure to be minimized is
Linear phase FIR Low-Pass eigenfilters
The solution Step1:Given ω p 、 ω s 、 α compute P Step2: Compute the eigenvalue and eigenvector of P Step3: Find smallest eigenvalue corresponding eigenvector
Example
The end