Back to results
US 2012/0059613 A1Published application
Signal parameter estimator
AssigneeThe Aerospace Corporation
InventorsAlexander C. Utter, Eugene Grayver
Life of this patent
Estimating the term…
Abstract
A signal parameter estimator includes curve fitting of multiple ideal spectral shapes.
Claims8
A method of characterizing an arbitrary radio frequency signal comprising the steps of:
- acquiring an input frequency spectrum;
- modeling at least a portion of the spectrum with a plurality of competing curve fits;
- determining a goodness of fit metric for each curve fit;
- and,
- selecting the best fit from among the plurality of curve fits.
2 more elements
A method of characterizing an arbitrary radio frequency signal comprising the steps of:
- providing a time domain signal and a first frequency spectrum representing the time domain signal;
- identifying a candidate signal in the spectrum and indicia of the candidate signal;
- transferring information derived from the candidate signal indicia to each of a downconverter, a low pass filter, and a decimator;
- downconverting the time domain signal in the downconverter to produce a downconverted signal;
- filtering the downconverted signal in the low pass filter to produce a filtered signal;
- decimating the filtered signal in the decimator to produce a decimated signal;
- providing a second frequency spectrum representing the decimated signal;
- fitting a plurality of shape equations to the second spectrum;
- calculating a goodness of fit metric for each fit;
- and,
- selecting the best fit and identifying a signal shape inherent in the second spectrum based on the goodness of fit metrics.
8 more elements
A method of characterizing an arbitrary radio frequency signal comprising the steps of:
- acquiring an input spectrum;
- from one or more signals in the spectrum, identifying a signal of interest (SOI);
- isolating the SOI;
- sampling SOI power spectral density as a function of a frequency derived variable;
- selecting a curve fit expression corresponding to a first ideal spectral shape;
- finding values of curve fit constants tending to minimize differences between the curve fit and the SOI samples;
- determining a figure of merit based on the differences between the curve fit and the SOI samples;
- repeating the selecting, finding, and determining steps for at least a second ideal spectral shape;
- and,
- characterizing the SOI by matching the SOI with the ideal spectral shape indicated by the figure of merit.
7 more elements
A method of characterizing an arbitrary radio frequency signal comprising the steps of:
- from an input spectrum, acquiring i samples of power spectral density (ISPSDi) as a function of a frequency derived variable (FDVi);
- selecting a curve fit expression (CFPSD=g(FDV)) corresponding to a first ideal spectral shape;
- finding values of curve fit constants tending to minimize differences between the curve fit and the input spectrum samples;
- determining a figure of merit based on the differences between the curve fit and the input spectrum samples;
- returning to the selecting step for at least a second ideal spectral shape;
- developing a first estimate of the input spectrum by matching the input spectrum with the ideal spectral shape indicated by the figure of merit;
- depleting the input spectrum by subtracting values of CFPSD evaluated at FDVi from corresponding values of ISPSDi;
- evaluating the reduced input spectrum for remaining signal content of interest;
- where signal content of interest remains, acquiring from the depleted input spectrum i samples of power spectral density (ISPSDi) as a function of a frequency derived variable (FDVi) and returning to the selecting step;
- and,
- characterizing the input spectrum with the collection of input spectrum estimates.
8 more elements
Dependencies are inferred from the claim text.