MSGR-based Low Latency Complex Matrix Inversion Architecture

L. Ma, K. Dickson, J. Mcallister, J. Mccanny

Research output: Chapter in Book/Report/Conference proceedingOther chapter contribution

3 Citations (Scopus)

Abstract

This paper presents a matrix inversion architecture based on the novel Modified Squared Givens Rotations (MSGR) algorithm, which extends the original SGR method to complex valued data, and also corrects erroneous results in the original SGR method when zeros occur on the diagonal of the matrix either initially or during processing. The MSGR algorithm also avoids complex dividers in the matrix inversion, thus minimising the complexity of potential real-time implementations. A systolic array architecture is implemented and FPGA synthesis results indicate a high-throughput low-latency complex matrix inversion solution. © 2008 IEEE.
Original languageEnglish
Title of host publicationInternational Conference on Signal Processing Proceedings, ICSP
Pages410-413
Number of pages4
DOIs
Publication statusPublished - Oct 2008

Fingerprint

Systolic arrays
Field programmable gate arrays (FPGA)
Throughput
Processing

Cite this

Ma, L., Dickson, K., Mcallister, J., & Mccanny, J. (2008). MSGR-based Low Latency Complex Matrix Inversion Architecture. In International Conference on Signal Processing Proceedings, ICSP (pp. 410-413) https://doi.org/10.1109/ICOSP.2008.4697158
Ma, L. ; Dickson, K. ; Mcallister, J. ; Mccanny, J. / MSGR-based Low Latency Complex Matrix Inversion Architecture. International Conference on Signal Processing Proceedings, ICSP. 2008. pp. 410-413
@inbook{1397e40eef9742e2906cc9afe8d3acf4,
title = "MSGR-based Low Latency Complex Matrix Inversion Architecture",
abstract = "This paper presents a matrix inversion architecture based on the novel Modified Squared Givens Rotations (MSGR) algorithm, which extends the original SGR method to complex valued data, and also corrects erroneous results in the original SGR method when zeros occur on the diagonal of the matrix either initially or during processing. The MSGR algorithm also avoids complex dividers in the matrix inversion, thus minimising the complexity of potential real-time implementations. A systolic array architecture is implemented and FPGA synthesis results indicate a high-throughput low-latency complex matrix inversion solution. {\circledC} 2008 IEEE.",
author = "L. Ma and K. Dickson and J. Mcallister and J. Mccanny",
year = "2008",
month = "10",
doi = "10.1109/ICOSP.2008.4697158",
language = "English",
pages = "410--413",
booktitle = "International Conference on Signal Processing Proceedings, ICSP",

}

Ma, L, Dickson, K, Mcallister, J & Mccanny, J 2008, MSGR-based Low Latency Complex Matrix Inversion Architecture. in International Conference on Signal Processing Proceedings, ICSP. pp. 410-413. https://doi.org/10.1109/ICOSP.2008.4697158

MSGR-based Low Latency Complex Matrix Inversion Architecture. / Ma, L.; Dickson, K.; Mcallister, J.; Mccanny, J.

International Conference on Signal Processing Proceedings, ICSP. 2008. p. 410-413.

Research output: Chapter in Book/Report/Conference proceedingOther chapter contribution

TY - CHAP

T1 - MSGR-based Low Latency Complex Matrix Inversion Architecture

AU - Ma, L.

AU - Dickson, K.

AU - Mcallister, J.

AU - Mccanny, J.

PY - 2008/10

Y1 - 2008/10

N2 - This paper presents a matrix inversion architecture based on the novel Modified Squared Givens Rotations (MSGR) algorithm, which extends the original SGR method to complex valued data, and also corrects erroneous results in the original SGR method when zeros occur on the diagonal of the matrix either initially or during processing. The MSGR algorithm also avoids complex dividers in the matrix inversion, thus minimising the complexity of potential real-time implementations. A systolic array architecture is implemented and FPGA synthesis results indicate a high-throughput low-latency complex matrix inversion solution. © 2008 IEEE.

AB - This paper presents a matrix inversion architecture based on the novel Modified Squared Givens Rotations (MSGR) algorithm, which extends the original SGR method to complex valued data, and also corrects erroneous results in the original SGR method when zeros occur on the diagonal of the matrix either initially or during processing. The MSGR algorithm also avoids complex dividers in the matrix inversion, thus minimising the complexity of potential real-time implementations. A systolic array architecture is implemented and FPGA synthesis results indicate a high-throughput low-latency complex matrix inversion solution. © 2008 IEEE.

UR - http://www.scopus.com/inward/record.url?partnerID=yv4JPVwI&eid=2-s2.0-67249157964&md5=af33776025d0649a4bae7a208e4856c3

U2 - 10.1109/ICOSP.2008.4697158

DO - 10.1109/ICOSP.2008.4697158

M3 - Other chapter contribution

SP - 410

EP - 413

BT - International Conference on Signal Processing Proceedings, ICSP

ER -

Ma L, Dickson K, Mcallister J, Mccanny J. MSGR-based Low Latency Complex Matrix Inversion Architecture. In International Conference on Signal Processing Proceedings, ICSP. 2008. p. 410-413 https://doi.org/10.1109/ICOSP.2008.4697158