A Bayesian approach to frequency line tracking
Abstract
A computationally efficient Bayesian frequency line tracker is presented. This frequency tracker has been implemented and operationally tested in the Navy's Passive Tracking Algorithm (PTA) program. Acoustic sensors receive narrowband signature from an ocean target. Using power spectral data as input, a discrete probability density function (PDF) of the received frequency is propagated in time using a first order Markov process to model line dynamics. The minimum mean squared error estimate (i.e., the posterior mean) and its standard deviation are obtained from the a posteriori PDF for use in the PTA Doppler target tracking algorithm. The frequency tracking algorithm is derived from a mathematical model, and results are presented.
 Publication:

EASCON '81; Electronics and Aerospace Systems Conventions
 Pub Date:
 1981
 Bibcode:
 1981easc.conf..286B
 Keywords:

 Bayes Theorem;
 Line Spectra;
 Radar Tracking;
 Run Time (Computers);
 Signal Processing;
 Tracking Problem;
 Algorithms;
 Markov Processes;
 Power Spectra;
 Probability Density Functions;
 RootMeanSquare Errors;
 Communications and Radar