Optimal kernels for time-frequency analysis

Richard G. Baraniuk, Douglas L. Jones

Research output: Contribution to journalConference article

Abstract

Current bilinear time-frequency representations apply a fixed kernel to smooth the Wigner distribution. However, the choice of a fixed kernel limits the class of signals that can be analyzed effectively. This paper presents optimality criteria for the design of signal-dependent kernels that suppress cross-components while passing as much auto-component energy as possible, irrespective of the form of the signal. A fast algorithm for the optimal kernel solution makes the procedure competitive computationally with fixed kernel methods. Examples demonstrate the superior performance of the optimal kernel for a frequency modulated signal.

Original languageEnglish (US)
Pages (from-to)181-187
Number of pages7
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume1348
DOIs
StatePublished - Nov 1 1990
EventAdvanced Signal Processing Algorithms, Architectures, and Implementations 1990 - San Diego, United States
Duration: Jul 8 1990Jul 13 1990

Fingerprint

Optimal Kernel
Time-frequency Analysis
kernel
Wigner Distribution
Kernel Methods
Optimality Criteria
Fast Algorithm
Dependent
Energy
Demonstrate
energy

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

Optimal kernels for time-frequency analysis. / Baraniuk, Richard G.; Jones, Douglas L.

In: Proceedings of SPIE - The International Society for Optical Engineering, Vol. 1348, 01.11.1990, p. 181-187.

Research output: Contribution to journalConference article

@article{3bec391bf307427aac814e06fd0c554a,
title = "Optimal kernels for time-frequency analysis",
abstract = "Current bilinear time-frequency representations apply a fixed kernel to smooth the Wigner distribution. However, the choice of a fixed kernel limits the class of signals that can be analyzed effectively. This paper presents optimality criteria for the design of signal-dependent kernels that suppress cross-components while passing as much auto-component energy as possible, irrespective of the form of the signal. A fast algorithm for the optimal kernel solution makes the procedure competitive computationally with fixed kernel methods. Examples demonstrate the superior performance of the optimal kernel for a frequency modulated signal.",
author = "Baraniuk, {Richard G.} and Jones, {Douglas L.}",
year = "1990",
month = "11",
day = "1",
doi = "10.1117/12.23475",
language = "English (US)",
volume = "1348",
pages = "181--187",
journal = "Proceedings of SPIE - The International Society for Optical Engineering",
issn = "0277-786X",
publisher = "SPIE",

}

TY - JOUR

T1 - Optimal kernels for time-frequency analysis

AU - Baraniuk, Richard G.

AU - Jones, Douglas L.

PY - 1990/11/1

Y1 - 1990/11/1

N2 - Current bilinear time-frequency representations apply a fixed kernel to smooth the Wigner distribution. However, the choice of a fixed kernel limits the class of signals that can be analyzed effectively. This paper presents optimality criteria for the design of signal-dependent kernels that suppress cross-components while passing as much auto-component energy as possible, irrespective of the form of the signal. A fast algorithm for the optimal kernel solution makes the procedure competitive computationally with fixed kernel methods. Examples demonstrate the superior performance of the optimal kernel for a frequency modulated signal.

AB - Current bilinear time-frequency representations apply a fixed kernel to smooth the Wigner distribution. However, the choice of a fixed kernel limits the class of signals that can be analyzed effectively. This paper presents optimality criteria for the design of signal-dependent kernels that suppress cross-components while passing as much auto-component energy as possible, irrespective of the form of the signal. A fast algorithm for the optimal kernel solution makes the procedure competitive computationally with fixed kernel methods. Examples demonstrate the superior performance of the optimal kernel for a frequency modulated signal.

UR - http://www.scopus.com/inward/record.url?scp=0025538867&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0025538867&partnerID=8YFLogxK

U2 - 10.1117/12.23475

DO - 10.1117/12.23475

M3 - Conference article

AN - SCOPUS:0025538867

VL - 1348

SP - 181

EP - 187

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

SN - 0277-786X

ER -