Impact of Parametric Uncertainties on Scattered Signal Distributions and Receiver Operating Characteristics

Abstract

Many different distributions are used to model statistics of waves that have been randomly scattered in atmospheric and terrain environments. These distributions have varying analytical advantages and ranges of physical applicability. This report reviews several basic distributions and discusses how they can be extended to include spatial and temporal variability in the scattering process. For this purpose, a compound probability density function (pdf) can be introduced in which a basic pdf describing the underlying scattering process is modulated by a second pdf describing parametric uncertainties in the scattering. We describe some useful new formulations based on the compound pdf, including strong and Rytov (lognormal) scattering processes modulated by the environment. These new formulations lead to relatively simple marginalized signal power distributions (Lomax and lognormal, respectively). Furthermore, we show how the conditional scattered signal pdf may be viewed as a likelihood function in which the modulating pdf is the Bayesian conjugate prior. The parameters of the modulating process can thus be refined by simple sequential Bayesian updating. Finally, the impact of the parametric uncertainties on signal detection and receiver operating characteristic curves is discussed and shown to be a very important consideration in practical applications.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 2018
Accession Number
AD1064130

Entities

People

  • Carl R Hart
  • Chris L. Pettit
  • D. K. Wilson
  • Daniel J. Breton
  • Edward T. Nykaza
  • Vladimir E. Ostashev

Tags

Communities of Interest

  • Energy and Power Technologies
  • Sensors

DTIC Thesaurus Topics

  • Acoustics
  • Automated Target Recognition
  • Computational Science
  • Data Science
  • Detection
  • Detectors
  • Engineering
  • Environment
  • Information Science
  • Monte Carlo Method
  • Probability
  • Probability Density Functions
  • Random Variables
  • Scattering
  • Signal Processing
  • Statistics
  • Target Recognition

Readers

  • Radar Systems Engineering.
  • Statistical inference.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference