Improved detection and tracking of dynamic signals by Bayes-Markov techniques

A recursive Bayesian technique is developed which computes the a posteriori probability density of the location of a dynamic signal in successive sets of low signal-to-noise ratio data. This serves to enhance the detection and tracking of signals which may be undetectable in an individual data frame and, because of unknown target motion between data frames, it may not generally be possible by conventional techniques to integrate the successive data for signal-to-noise enhancement. Computer simulation examples are presented to demonstrate the performance of the Bayesian-Markov technique on simulated low SNR range-doppler amplitude data obtained in a pulse-doppler radar system and on simulated cross-ambiguity surface data obtained by cross correlating the data received at two spatially separated sonar arrays.