Model Order Determination Methods for Autoregressive Radio Direction Finding Techniques
Abstract
This report reviews methods to determine the optimum model order of adaptive filter techniques applied to the problem of tactical radio direction finding (DF), and the related issue of the determination of the number of signals for adaptive filter techniques using eigen decomposition techniques. This work was carried out in support of research into advanced radio direction finding techniques for land tactical applications. For radio direction finding applications, adjusting the adaptive filter model order to a value less than the maximum (i.e., one less than the number of antennas in the sensor array) increases the ability of an adaptive filter based DF algorithm to resolve multipath signals. It is from this point of view that model order determination methods are discussed in this report. In general, the optimum model order is not known beforehand, so a number of model order determination methods have been proposed to tackle this problem. In this report, these are classified according to the autocorrelation matrix estimator used. Three classes of estimators are considered, which are: (a) the autocorrelation method, (b) the covariance and modified covariance method, and (c) the eigen decomposition method. (kr)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1990
- Accession Number
- ADA220707
Entities
People
- W. Read
Organizations
- Defence Research and Development Canada