Two Dimensional Uncorrelated Bias in Fix Algorithms.

Abstract

Bias in Location Estimation Algorithms is defined to be a vector from the estimated fix site to the true Fix site. Four Fix Algorithms are examined in this report: Least Squares (Stanfield Approximation), Weighted Least Squares, Minimization of Angular Error, and minimization of the Site of the Angular Error. Analysis expressions for the First-Order error are derived for two specific cases: Two Lines-of-Bearing (LOB's) and infinite LOB's. The analytic derivations are relegated to appendices. Three questions are answered by this report. First, which algorithms cause the largest bias? Second, how sensitive is this bias to sample size, system noise, and variance? Third, which algorithms (if any) are asymptotically unbiased? All conclusions are then verified by simulation. Keywords: Bias; Least squares; Minimization of angular error; Asymptotically unbiased; Simulation; Weights; Geometric effects.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 10, 1987
Accession Number
ADA189473

Entities

People

  • Michael W. Rennie

Organizations

  • Jet Propulsion Laboratory

Tags

Communities of Interest

  • Sensors

DTIC Thesaurus Topics

  • Algorithms
  • Army Intelligence
  • California
  • Errors
  • Jet Propulsion
  • Least Squares Method
  • Mathematical Analysis
  • Security
  • Simulations
  • Standards
  • Symmetry
  • Two Dimensional
  • United States

Readers

  • Approximation Theory.