Discrimination of boosted trajectories among several radar observed objects

In the present paper an algorithm is proposed for the identification of a missile trajectory and impact point prediction on the basis of radar measurements. An analytic solution of a simplified missile dynamic model is applied to evaluate the missile parameters and state variables in boost phase. These analytic estimates are used as guess values to initialize a Maximum Likelihood. Moreover an Extended Kalman Filter is used to investigate the possibility of a real time detection algorithm. It is shown that the filter may have problematic convergence in the boost phase without sufficiently accurate preliminary information on the missile. On the other hand, Extended Kalman Filter is shown to be effective in the ballistic and reentry phase, so a blend of the Maximum Likelihood Estimate and Extended Kalman Filter is proposed. Furthermore the Kalman filter allows to propagate in real time the dynamic model, evaluating at the same time discrepancies between actual state variables measurements and dynamical model propagation. The algorithm analyzes dynamical missile model and effective measurements to discriminate between radar false alarms and real targets. The analysis of filter innovations standard deviation provides a chance to discriminate between real targets and false alarms on the basis of different area versus mass ratios. Thus, in a multitarget environments, a filters bank permits to achieve further information on target nature beside radar echoes. Simulations of possible real targets and false alarms have been performed and discrimination results are discussed.