Academic Journals Database
Disseminating quality controlled scientific knowledge

Reducting Power Dissipation in Fir Filter: an Analysis

ADD TO MY LIST
 
Author(s): Rakesh Kumar Bansal | Manoj Garg | Savina Bansal

Journal: Signal Processing : An International Journal
ISSN 1985-2339

Volume: 4;
Issue: 1;
Start page: 62;
Date: 2010;
Original page

Keywords: FIR | Coefficient segmentation | Steepest decent | SPTl | Power reduction

ABSTRACT
In this paper, three existing techniques, Signed Power-of-Two (SPT), Steepest decent and Coefficient segmentation, for power reduction of FIR filters are analyzed. These techniques reduce switching activity which is directly related to the power consumption of a circuit. In an FIR filter, the multiplier consumes maximum power. Therefore, power consumption can be reduced either by by making the filter multiplier-less or by minimizing hamming distance between the coefficients of this multiplier as it directly translates into reduction in power dissipation [8]. The results obtained on four filters (LP) show that hamming distance can be reduced upto 26% and 47% in steepest decent and coefficient segmentation algorithm respectively. Multiplier-less filter can be realized by realizing coefficients in signed power-of-two terms, i.e. by shifting and adding the coefficients, though at the cost of shift operation overhead.

Tango Jona
Tangokurs Rapperswil-Jona

     Save time & money - Smart Internet Solutions