The Application of Linear Prediction to Sequential Classification of Radar Target Signatures

Abstract

A linear-predictive form of the quadratic classifier (the optimal decision rule for Gaussian random processes) is developed and applied to the discrimination and classification of radar target signatures. The classifier was devised to implement a sequential probability ratio test (SPRT); that is, consecutive radar returns are observed until the target can be classified with a prescribed probability of error. Because of the linear-predictive formulation, the computational and storage requirements for the classifier are related only to the number of returns necessary to predict the signature and not to the length of signature observed; a classifier with modest storage and computational requirements can be employed to classify signatures consisting of an arbitrarily large number of radar returns. The classifier is related to several results in mean-square filtering theory and has an interpretation in terms of the maximum entropy and maximum likelihood spectral estimates for the target signatures.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 25, 1976
Accession Number
ADA026695

Entities

People

  • Charles W. Therrien

Organizations

  • Massachusetts Institute of Technology

Tags

Communities of Interest

  • Energy and Power Technologies
  • Human Systems
  • Sensors
  • Space

DTIC Thesaurus Topics

  • Ballistic Missiles
  • Computations
  • Covariance
  • Data Science
  • Detection
  • Frequency
  • Frequency Domain
  • Gaussian Noise
  • Gaussian Processes
  • Information Science
  • Noise
  • Probability
  • Radar Signatures
  • Radar Targets
  • Regression Analysis
  • Statistics
  • Target Signatures

Readers

  • Computer Vision.
  • Radar Systems Engineering.
  • Statistical inference.