Instantaneous Power Spectrum

Abstract

The estimation of time varying spectra is a complicated one. The use of classical techniques coupled with the local stationarity assumption is met with only moderate success. Of the many time frequency distribution functions used in the signal analysis, none present fully satisfactory spectra. The performance of the spectrogram, Instantaneous Power Spectra (IPS) the Wigner- Ville Distribution (WD) and various aspects of the Rihaczek distribution (RD) for a variety of signal nonstationarities are compared. WD has the most narrow main-lobes but suffers from spectral cross-terms. IPS, the real part of the RD consistently shows a broadened main-lobe without cross-terms. The squared magnitude of the RD places sharp peaks along the crest of the main-lobe and is otherwise very similar to IPS. The imaginary part of the RD shows a sensitivity to discontinuous frequency changes i.e., frequency shift keying. Keywords: Theses, Instantaneous Power Spectrum, spectral estimation, nonstationationary signal analysis.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1990
Accession Number
ADA229098

Entities

People

  • Elizabeth H. Stitz

Organizations

  • Naval Postgraduate School

Tags

Communities of Interest

  • Air Platforms
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Difference Frequency
  • Distribution Functions
  • Electrical Engineering
  • Engineering
  • Estimators
  • Frequency
  • Frequency Shift
  • Information Theory
  • Kernel Functions
  • Pattern Recognition
  • Power Spectra
  • Random Variables
  • Signal Processing
  • Spectra
  • Stochastic Processes

Fields of Study

  • Engineering

Readers

  • Phased Array Antenna Design.
  • Radio communications and signal processing.
  • Statistical inference.