Fast and energy-efficient constant-coefficient FIR filters using residue number system

Piotr Patronik, Krzysztof Berezowski, Stanislaw J. Piestrak, Janusz Biernat, Aviral Shrivastava

Research output: Chapter in Book/Report/Conference proceedingConference contribution

12 Citations (Scopus)

Abstract

In this paper, we present constant-coefficient finite impulse response (FIR) filters design using residue number system (RNS) arithmetic. The novelty of our approach rests in an attempt to maximize the accumulated benefit of the application of RNS to the design of constant coefficient filters. To achieve this, we consider the impact of RNS on many layers: from coefficient representation and techniques of sharing of subexpressions in the multiplier block (MB), to its optimized usage in the MB and accumulation pipeline hardware design. As a result, we propose a common subexpression elimination (CSE) based synthesis technique for RNS-based MBs, along with a high-performance RNS-based FIR filter architecture that employs RNS arithmetic principles but implements them mainly using more efficient 2's complement hardware. Several filters with numbers of taps ranging from 25 to 326 and dynamic ranges from 24 to 50 bits have been synthesized using TSMC 90 nm LP kit and Cadence RTL Compiler. Comparison of power, delay, and area of the new filters implemented using the 4- and 5-moduli RNSs against various equivalent 2's complement counterparts show uniform improvement in performance and power efficiency, often accompanied by significant reduction in area/power consumption as compared to 2's complement implementations. We observed up to 22% improvement in peformance (19% reduction in area) within bounded power envelope, or up to 14% reduction in power consumption (12% reduction in area) at same frequency.

Original languageEnglish (US)
Title of host publicationProceedings of the International Symposium on Low Power Electronics and Design
Pages385-390
Number of pages6
DOIs
StatePublished - 2011
Event17th IEEE/ACM International Symposium on Low Power Electronics and Design, ISLPED 2011 - Fukuoka, Japan
Duration: Aug 1 2011Aug 3 2011

Other

Other17th IEEE/ACM International Symposium on Low Power Electronics and Design, ISLPED 2011
CountryJapan
CityFukuoka
Period8/1/118/3/11

Fingerprint

Numbering systems
FIR filters
Electric power utilization
Hardware
Pipelines

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Patronik, P., Berezowski, K., Piestrak, S. J., Biernat, J., & Shrivastava, A. (2011). Fast and energy-efficient constant-coefficient FIR filters using residue number system. In Proceedings of the International Symposium on Low Power Electronics and Design (pp. 385-390). [5993671] https://doi.org/10.1109/ISLPED.2011.5993671

Fast and energy-efficient constant-coefficient FIR filters using residue number system. / Patronik, Piotr; Berezowski, Krzysztof; Piestrak, Stanislaw J.; Biernat, Janusz; Shrivastava, Aviral.

Proceedings of the International Symposium on Low Power Electronics and Design. 2011. p. 385-390 5993671.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Patronik, P, Berezowski, K, Piestrak, SJ, Biernat, J & Shrivastava, A 2011, Fast and energy-efficient constant-coefficient FIR filters using residue number system. in Proceedings of the International Symposium on Low Power Electronics and Design., 5993671, pp. 385-390, 17th IEEE/ACM International Symposium on Low Power Electronics and Design, ISLPED 2011, Fukuoka, Japan, 8/1/11. https://doi.org/10.1109/ISLPED.2011.5993671
Patronik P, Berezowski K, Piestrak SJ, Biernat J, Shrivastava A. Fast and energy-efficient constant-coefficient FIR filters using residue number system. In Proceedings of the International Symposium on Low Power Electronics and Design. 2011. p. 385-390. 5993671 https://doi.org/10.1109/ISLPED.2011.5993671
Patronik, Piotr ; Berezowski, Krzysztof ; Piestrak, Stanislaw J. ; Biernat, Janusz ; Shrivastava, Aviral. / Fast and energy-efficient constant-coefficient FIR filters using residue number system. Proceedings of the International Symposium on Low Power Electronics and Design. 2011. pp. 385-390
@inproceedings{5c7c78d698d243cfbe06aec9200b7f5c,
title = "Fast and energy-efficient constant-coefficient FIR filters using residue number system",
abstract = "In this paper, we present constant-coefficient finite impulse response (FIR) filters design using residue number system (RNS) arithmetic. The novelty of our approach rests in an attempt to maximize the accumulated benefit of the application of RNS to the design of constant coefficient filters. To achieve this, we consider the impact of RNS on many layers: from coefficient representation and techniques of sharing of subexpressions in the multiplier block (MB), to its optimized usage in the MB and accumulation pipeline hardware design. As a result, we propose a common subexpression elimination (CSE) based synthesis technique for RNS-based MBs, along with a high-performance RNS-based FIR filter architecture that employs RNS arithmetic principles but implements them mainly using more efficient 2's complement hardware. Several filters with numbers of taps ranging from 25 to 326 and dynamic ranges from 24 to 50 bits have been synthesized using TSMC 90 nm LP kit and Cadence RTL Compiler. Comparison of power, delay, and area of the new filters implemented using the 4- and 5-moduli RNSs against various equivalent 2's complement counterparts show uniform improvement in performance and power efficiency, often accompanied by significant reduction in area/power consumption as compared to 2's complement implementations. We observed up to 22{\%} improvement in peformance (19{\%} reduction in area) within bounded power envelope, or up to 14{\%} reduction in power consumption (12{\%} reduction in area) at same frequency.",
author = "Piotr Patronik and Krzysztof Berezowski and Piestrak, {Stanislaw J.} and Janusz Biernat and Aviral Shrivastava",
year = "2011",
doi = "10.1109/ISLPED.2011.5993671",
language = "English (US)",
isbn = "9781612846590",
pages = "385--390",
booktitle = "Proceedings of the International Symposium on Low Power Electronics and Design",

}

TY - GEN

T1 - Fast and energy-efficient constant-coefficient FIR filters using residue number system

AU - Patronik, Piotr

AU - Berezowski, Krzysztof

AU - Piestrak, Stanislaw J.

AU - Biernat, Janusz

AU - Shrivastava, Aviral

PY - 2011

Y1 - 2011

N2 - In this paper, we present constant-coefficient finite impulse response (FIR) filters design using residue number system (RNS) arithmetic. The novelty of our approach rests in an attempt to maximize the accumulated benefit of the application of RNS to the design of constant coefficient filters. To achieve this, we consider the impact of RNS on many layers: from coefficient representation and techniques of sharing of subexpressions in the multiplier block (MB), to its optimized usage in the MB and accumulation pipeline hardware design. As a result, we propose a common subexpression elimination (CSE) based synthesis technique for RNS-based MBs, along with a high-performance RNS-based FIR filter architecture that employs RNS arithmetic principles but implements them mainly using more efficient 2's complement hardware. Several filters with numbers of taps ranging from 25 to 326 and dynamic ranges from 24 to 50 bits have been synthesized using TSMC 90 nm LP kit and Cadence RTL Compiler. Comparison of power, delay, and area of the new filters implemented using the 4- and 5-moduli RNSs against various equivalent 2's complement counterparts show uniform improvement in performance and power efficiency, often accompanied by significant reduction in area/power consumption as compared to 2's complement implementations. We observed up to 22% improvement in peformance (19% reduction in area) within bounded power envelope, or up to 14% reduction in power consumption (12% reduction in area) at same frequency.

AB - In this paper, we present constant-coefficient finite impulse response (FIR) filters design using residue number system (RNS) arithmetic. The novelty of our approach rests in an attempt to maximize the accumulated benefit of the application of RNS to the design of constant coefficient filters. To achieve this, we consider the impact of RNS on many layers: from coefficient representation and techniques of sharing of subexpressions in the multiplier block (MB), to its optimized usage in the MB and accumulation pipeline hardware design. As a result, we propose a common subexpression elimination (CSE) based synthesis technique for RNS-based MBs, along with a high-performance RNS-based FIR filter architecture that employs RNS arithmetic principles but implements them mainly using more efficient 2's complement hardware. Several filters with numbers of taps ranging from 25 to 326 and dynamic ranges from 24 to 50 bits have been synthesized using TSMC 90 nm LP kit and Cadence RTL Compiler. Comparison of power, delay, and area of the new filters implemented using the 4- and 5-moduli RNSs against various equivalent 2's complement counterparts show uniform improvement in performance and power efficiency, often accompanied by significant reduction in area/power consumption as compared to 2's complement implementations. We observed up to 22% improvement in peformance (19% reduction in area) within bounded power envelope, or up to 14% reduction in power consumption (12% reduction in area) at same frequency.

UR - http://www.scopus.com/inward/record.url?scp=80052773744&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80052773744&partnerID=8YFLogxK

U2 - 10.1109/ISLPED.2011.5993671

DO - 10.1109/ISLPED.2011.5993671

M3 - Conference contribution

AN - SCOPUS:80052773744

SN - 9781612846590

SP - 385

EP - 390

BT - Proceedings of the International Symposium on Low Power Electronics and Design

ER -