Frequency and Phase Break Detection

Abstract

A phase break detection method is proposed and successfully tested for the CANVAS analysis software for frequency standards. Phase breaks are detected above the noise floor by classifying a very large phase step as a phase break. This detection method succeeds in numerically quantifying a large visual jump in the phase plot without using input parameters other than the phase data. These phase breaks are automatically identified by the software, and then are removed from the data. To implement this solution, it is assumed that the frequency standard is well behaved. The most extreme phase steps (1% of the total data) are assumed to contain all phase breaks and other misbehaving data points, and this small subset of 1% is neglected during the detection method's preliminary analysis. If these assumptions are violated, then this phase-break detection method does not apply to the set. The phase-break detection algorithm still needs to be interfaced with the CANVAS user interface. Also, this method and the frequency-break detection method are intended for post-process use. A frequency-break detection method is also proposed, and the assumptions that invalidate the method are explained.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 2009
Accession Number
ADA518031

Entities

People

  • Scott Czopek

Organizations

  • University of Colorado Boulder

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Algorithms
  • Atomic Beam Masers
  • Atomic Clocks
  • Clocks
  • Data Sets
  • Detection
  • Frequency
  • Frequency Standards
  • Information Science
  • Intervals
  • Life Tests
  • Masers
  • Military Research
  • Normal Distribution
  • Phase
  • Standards
  • Time Intervals

Readers

  • Applied Combinatorial Optimization and Logic Circuit Design.
  • Radar Systems Engineering.
  • Regression Analysis.