ABSTRACT
This chapter considers the basic problem of passive localization and motion analysis of a source observed from a linear array. The estimator is an expanding memory filter that uses successive time delay observations together with an appropriate physical model of source motion to obtain smooth source trajectory estimates. Estimation of contact course and speed is only fruitful using a filtering process. Mean errors in the Kalman filter localization and motion solutions obtained by processing, first, bearing and range measurements, then, time delays, are displayed. A modified polar Kalman filter is used to avoid the Cartesian filter problems triggered by the lack of observability prior to observer maneuvers. The choice of an appropriate processing structure is critical to successful estimation. The right structure improves substantially on the techniques that process inappropriately mapped time delays or those that transform directly the best time delays available into source motion estimates.
