Performance Comparison of a Sub-Optimal Array Processor.

Abstract

The purpose of the research is to compare the performance of two detection receivers for the signal known except for direction (SKED) array problem. Each receiver represents a different approach to the problem's solution--one is designed under the Bayesian philosophy (parametric) and the other uses a classical beam sweeping approach. These are noted as the optimal and sub-optimal detectors, respectively. It is shown that both processors can be implemented as estimate and plug structures using the conditional likelihood ratio. The two estimators used are the pseudo estimate (X(PSE)) and maximum likelihood estimate (X(MLE)). An in depth discussion of these two estimators reveals that X(PSE) is simply equal to X(MLE) plus a data dependent correction term. A performance comparison is made for the two and three element array cases. It is shown that the sub-optimal detector performs the same as the optimal detector when A sub zero = zero (uniform a priori knowledge). Furthermore, for a two element array, the results are proven analytically. (Author)

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1972
Accession Number
AD0762048

Entities

People

  • William Searles Hodgkiss Jr

Organizations

  • Duke University

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Detection
  • Detectors
  • Estimators
  • Philosophy
  • Warning Systems

Fields of Study

  • Engineering

Readers

  • Data Mining and Knowledge Discovery.
  • Educational Psychology
  • Statistical inference.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference