Finite-precision error analysis of QRD-RLS and STAR-RLS adaptive filters

Kalavai J. Raghunath, Keshab K Parhi

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

The QR decomposition-based recursive least-squares (RLS) adaptive filtering (QRD-RLS) algorithm is suitable for VLSI implementation since it has good numerical properties and can be mapped onto a systolic array. Recently, a new fine-grain pipelinable STAR-RLS algorithm was developed. The pipelined STAR-RLS algorithm (PSTAR-RLS) is useful for high-speed applications. The stability of QRD-RLS, STAR-RLS, and PSTARRLS has been proved, but the performance of these algorithms in finite-precision arithmetic has not yet been analyzed. The aim of this paper is to determine expressions for the degradation in the performance of these algorithms due to finite precision. By exploiting the steady-state properties of these algorithms, simple expressions are obtained that depend only on known parameters. This analysis can be used to compare the algorithms and to decide the wordlength to be used in an implementation. Since floating- or fixed-point arithmetic representations may be used in practice, both representations are considered in this paper. The results show that the three algorithms have about the same finiteprecision performance, with PSTAR-RLS performing better than STAR-RLS, which does better than QRD-RLS. These algorithms can be implemented with as few as 8 bits for the fractional part, depending on the filter size and the forgetting factor used. The theoretical expressions are found to be in good agreement with the simulation results.

Original languageEnglish (US)
Pages (from-to)1193-1209
Number of pages17
JournalIEEE Transactions on Signal Processing
Volume45
Issue number5
DOIs
StatePublished - Dec 1 1997

Fingerprint

Adaptive filters
Error analysis
Fixed point arithmetic
Digital arithmetic
Systolic arrays
Adaptive filtering
Decomposition
Degradation

Cite this

Finite-precision error analysis of QRD-RLS and STAR-RLS adaptive filters. / Raghunath, Kalavai J.; Parhi, Keshab K.

In: IEEE Transactions on Signal Processing, Vol. 45, No. 5, 01.12.1997, p. 1193-1209.

Research output: Contribution to journalArticle

@article{ae026958cfb14cd4bd5ec9d0e708df83,
title = "Finite-precision error analysis of QRD-RLS and STAR-RLS adaptive filters",
abstract = "The QR decomposition-based recursive least-squares (RLS) adaptive filtering (QRD-RLS) algorithm is suitable for VLSI implementation since it has good numerical properties and can be mapped onto a systolic array. Recently, a new fine-grain pipelinable STAR-RLS algorithm was developed. The pipelined STAR-RLS algorithm (PSTAR-RLS) is useful for high-speed applications. The stability of QRD-RLS, STAR-RLS, and PSTARRLS has been proved, but the performance of these algorithms in finite-precision arithmetic has not yet been analyzed. The aim of this paper is to determine expressions for the degradation in the performance of these algorithms due to finite precision. By exploiting the steady-state properties of these algorithms, simple expressions are obtained that depend only on known parameters. This analysis can be used to compare the algorithms and to decide the wordlength to be used in an implementation. Since floating- or fixed-point arithmetic representations may be used in practice, both representations are considered in this paper. The results show that the three algorithms have about the same finiteprecision performance, with PSTAR-RLS performing better than STAR-RLS, which does better than QRD-RLS. These algorithms can be implemented with as few as 8 bits for the fractional part, depending on the filter size and the forgetting factor used. The theoretical expressions are found to be in good agreement with the simulation results.",
author = "Raghunath, {Kalavai J.} and Parhi, {Keshab K}",
year = "1997",
month = "12",
day = "1",
doi = "10.1109/78.575694",
language = "English (US)",
volume = "45",
pages = "1193--1209",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "5",

}

TY - JOUR

T1 - Finite-precision error analysis of QRD-RLS and STAR-RLS adaptive filters

AU - Raghunath, Kalavai J.

AU - Parhi, Keshab K

PY - 1997/12/1

Y1 - 1997/12/1

N2 - The QR decomposition-based recursive least-squares (RLS) adaptive filtering (QRD-RLS) algorithm is suitable for VLSI implementation since it has good numerical properties and can be mapped onto a systolic array. Recently, a new fine-grain pipelinable STAR-RLS algorithm was developed. The pipelined STAR-RLS algorithm (PSTAR-RLS) is useful for high-speed applications. The stability of QRD-RLS, STAR-RLS, and PSTARRLS has been proved, but the performance of these algorithms in finite-precision arithmetic has not yet been analyzed. The aim of this paper is to determine expressions for the degradation in the performance of these algorithms due to finite precision. By exploiting the steady-state properties of these algorithms, simple expressions are obtained that depend only on known parameters. This analysis can be used to compare the algorithms and to decide the wordlength to be used in an implementation. Since floating- or fixed-point arithmetic representations may be used in practice, both representations are considered in this paper. The results show that the three algorithms have about the same finiteprecision performance, with PSTAR-RLS performing better than STAR-RLS, which does better than QRD-RLS. These algorithms can be implemented with as few as 8 bits for the fractional part, depending on the filter size and the forgetting factor used. The theoretical expressions are found to be in good agreement with the simulation results.

AB - The QR decomposition-based recursive least-squares (RLS) adaptive filtering (QRD-RLS) algorithm is suitable for VLSI implementation since it has good numerical properties and can be mapped onto a systolic array. Recently, a new fine-grain pipelinable STAR-RLS algorithm was developed. The pipelined STAR-RLS algorithm (PSTAR-RLS) is useful for high-speed applications. The stability of QRD-RLS, STAR-RLS, and PSTARRLS has been proved, but the performance of these algorithms in finite-precision arithmetic has not yet been analyzed. The aim of this paper is to determine expressions for the degradation in the performance of these algorithms due to finite precision. By exploiting the steady-state properties of these algorithms, simple expressions are obtained that depend only on known parameters. This analysis can be used to compare the algorithms and to decide the wordlength to be used in an implementation. Since floating- or fixed-point arithmetic representations may be used in practice, both representations are considered in this paper. The results show that the three algorithms have about the same finiteprecision performance, with PSTAR-RLS performing better than STAR-RLS, which does better than QRD-RLS. These algorithms can be implemented with as few as 8 bits for the fractional part, depending on the filter size and the forgetting factor used. The theoretical expressions are found to be in good agreement with the simulation results.

UR - http://www.scopus.com/inward/record.url?scp=0031142182&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031142182&partnerID=8YFLogxK

U2 - 10.1109/78.575694

DO - 10.1109/78.575694

M3 - Article

VL - 45

SP - 1193

EP - 1209

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

IS - 5

ER -