Stein's Method and Its Application in Radar Signal Processing

Abstract

In the analysis of radar signal processing, many approximations are made in order to simplify systems analysis. In the context of radar detection, environmental assumptions, such as in the modelling of clutter and noise, are often made. Similar assumptions are employed in the analysis of speckle in radar imaging systems. In addition to this, further statistical approximations of random processes of interest are made. In radar detection theory, an assumption on the distribution of a clutter statistic is often made. An important part of the model validation process is to assess the validity of employed assumptions. The validity of environmental assumptions is often assessed on the basis of field trials. In the case of approximations of random processes of interest, there are a number approaches. This study introduces a general scheme, known as Stein's Method, for assessing distributional approximations of random processes. It is described in general terms, and its application to the Poisson and Gaussian approximation of a given random variable is outlined. A new development of Stein's Method for Exponential approximation is included. The importance of the latter is that it can be used as na approximating distribution for a number of random variables of interest in radar.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 2005
Accession Number
ADA437309

Entities

People

  • Graham V. Weinberg

Organizations

  • Defence Science and Technology Group

Tags

Communities of Interest

  • Electronic Warfare
  • Space
  • Weapons Technologies

DTIC Thesaurus Topics

  • Computational Science
  • Detection
  • Detectors
  • Differential Equations
  • Distribution Functions
  • Electronic Warfare
  • Markov Processes
  • Radar
  • Radar Signals
  • Random Variables
  • Real Numbers
  • Signal Processing
  • Standards
  • Stochastic Processes
  • Systems Science
  • Warfare
  • Warning Systems

Readers

  • Radar Systems Engineering.
  • Statistical inference.