Effect of glint errors on steady-state tracking accuracies of maneuvering targets

Abstract

The effect of glint errors on target tracking accuracies is studied by various analytical means. The glint error is modeled as a first-order Markov process, characterized by a correlation time ( tau ) and a position error variance. The target motion is described by one of two dynamic models: the exponentially correlated velocity model and the exponentially correlated acceleration model. For both models, the steady-state variance of the estimated target velocity exhibits resonant behavior whenever the time-constant for target maneuvers is close to tau . The analysis is simplified through the use of some mathematical tricks, which may themselves be of interest to Kalman filter specialists.<>