A maximum likelihood tracker

Abstract : This report deals with the development and utilization of a maximum likelihood tracking algorithm designed to handle a single diffusion target. The tracker is required to accept or reject each of a sequence of discrete position reports, some of which are false alarms. Average tracking time, E(T), is defined and used to evaluate tracker performance. The effects of tracker memory are examined and a Kalman filter is developed to handle the effects of measurement error. Simulation results are included.