Microkicking for Fast Convergence of Sparse Kaczmarz and Sparse LMS

Michael Lunglmayr; Mario Huemer

doi:10.1109/CAMSAP.2017.8313085

Microkicking for Fast Convergence of Sparse Kaczmarz and Sparse LMS

Institute of Signal Processing

Research output: Chapter in Book/Report/Conference proceeding › Conference proceedings › peer-review

Abstract

Algorithms based on linearized Bregman iterations are able to perform sparse reconstruction at a low computational complexity. Especially the Least-Mean-Squares (LMS) and Kaczmarz variants of linearized Bregman iterations proved to be very feasible for fixed-point digital hardware implementation. We present a method that we call microkicking for improving the convergence speed of linearized Bregman based algorithms. This method can be implemented with only a negligible complexity overhead leading to significantly faster convergence for both variants of the linearized Bregman iterations. We furthermore show simulation results demonstrating the performance gains achievable by microkicking.

Original language	English
Title of host publication	Proceedings of the IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP 2017)
Publisher	IEEE
Number of pages	5
ISBN (Print)	978-1-5386-1251-4
DOIs	https://doi.org/10.1109/CAMSAP.2017.8313085
Publication status	Published - Dec 2017

Fields of science

202 Electrical Engineering, Electronics, Information Engineering
202015 Electronics
202022 Information technology
202027 Mechatronics
202037 Signal processing
202041 Computer engineering

JKU Focus areas

Computation in Informatics and Mathematics
Mechatronics and Information Processing

Access to Document

10.1109/CAMSAP.2017.8313085

Cite this

@inproceedings{d533ee4399474785a63a6717c1ccbdee,

title = "Microkicking for Fast Convergence of Sparse Kaczmarz and Sparse LMS",

abstract = "Algorithms based on linearized Bregman iterations are able to perform sparse reconstruction at a low computational complexity. Especially the Least-Mean-Squares (LMS) and Kaczmarz variants of linearized Bregman iterations proved to be very feasible for fixed-point digital hardware implementation. We present a method that we call microkicking for improving the convergence speed of linearized Bregman based algorithms. This method can be implemented with only a negligible complexity overhead leading to significantly faster convergence for both variants of the linearized Bregman iterations. We furthermore show simulation results demonstrating the performance gains achievable by microkicking.",

author = "Michael Lunglmayr and Mario Huemer",

year = "2017",

month = dec,

doi = "10.1109/CAMSAP.2017.8313085",

language = "English",

isbn = "978-1-5386-1251-4",

booktitle = "Proceedings of the IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP 2017)",

publisher = "IEEE",

}

TY - GEN

T1 - Microkicking for Fast Convergence of Sparse Kaczmarz and Sparse LMS

AU - Lunglmayr, Michael

AU - Huemer, Mario

PY - 2017/12

Y1 - 2017/12

N2 - Algorithms based on linearized Bregman iterations are able to perform sparse reconstruction at a low computational complexity. Especially the Least-Mean-Squares (LMS) and Kaczmarz variants of linearized Bregman iterations proved to be very feasible for fixed-point digital hardware implementation. We present a method that we call microkicking for improving the convergence speed of linearized Bregman based algorithms. This method can be implemented with only a negligible complexity overhead leading to significantly faster convergence for both variants of the linearized Bregman iterations. We furthermore show simulation results demonstrating the performance gains achievable by microkicking.

AB - Algorithms based on linearized Bregman iterations are able to perform sparse reconstruction at a low computational complexity. Especially the Least-Mean-Squares (LMS) and Kaczmarz variants of linearized Bregman iterations proved to be very feasible for fixed-point digital hardware implementation. We present a method that we call microkicking for improving the convergence speed of linearized Bregman based algorithms. This method can be implemented with only a negligible complexity overhead leading to significantly faster convergence for both variants of the linearized Bregman iterations. We furthermore show simulation results demonstrating the performance gains achievable by microkicking.

UR - https://ieeexplore.ieee.org/document/8313085

U2 - 10.1109/CAMSAP.2017.8313085

DO - 10.1109/CAMSAP.2017.8313085

M3 - Conference proceedings

SN - 978-1-5386-1251-4

BT - Proceedings of the IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP 2017)

PB - IEEE

ER -

Microkicking for Fast Convergence of Sparse Kaczmarz and Sparse LMS

Abstract

Fields of science

JKU Focus areas

Access to Document

Other files and links

Low Complexity Iterative Signal Processing Methods

Cite this

Microkicking for Fast Convergence of Sparse Kaczmarz and Sparse LMS

Abstract

Fields of science

JKU Focus areas

Access to Document

Other files and links

Projects

Low Complexity Iterative Signal Processing Methods

Cite this