Theory of Roundoff Noise in Cascade Realizations of Finite Impulse Response Digital Filters
David S. K. Chan, L. R. Rabiner
Abstract
David S. K. Chan, L. R. Rabiner
Abstract
This paper presents a theoretical treatment of the roundoff noise problem for the special case of cascade realizations of Finite Impulse Response (FIR) digital filters.†Explicit relations for evaluating roundoff noise with the usual assumption of uncorrelated samples are presented. Useful scaling methods are stated and classified as to conditions when these methods are optimum. Important differences between use of these scaling procedures for Infinite Impulse Response (IIR) filters and FIR filters are pointed out. Finally, useful properties of linear phase FIR filters are discussed.
OpenAlex reports 29 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper presents a theoretical treatment of the roundoff noise problem for the special case of cascade realizations of Finite Impulse Response (FIR) digital filters.†Explicit relations for evaluating roundoff noise with the usual assumption of uncorrelated samples are presented. Useful scaling methods are stated and classified as to conditions when these methods are optimum. Important differences between use of these scaling procedures for Infinite Impulse Response (IIR) filters and FIR filters are pointed out. Finally, useful properties of linear phase FIR filters are discussed.
Key concepts: Infinite impulse response, Finite impulse response, Digital filter, Linear filter, Cascade, Mathematics, Scaling, Uncorrelated