Azimuth-only localization and accuracy study for piecewise curvilinearly moving targets

Passive tracking of maneuvering emitters is of fundamental interest. For this purpose, an appropriate motion model, the piecewise curvilinear motion model, is presented in this paper. The high-dimensional target state is characterized by position, speed, course, piecewise constant tangential and normal velocity, as well as maneuver change-over times. A single moving sensor collects azimuth measurements to obtain the target state. In order to determine the maximum achievable estimation accuracy, we derive the Cramér-Rao bounds for this estimation problem. A maximum likelihood estimator is proposed to solve the localization problem and to calculate the maneuver change-over times. Estimation results are obtained in Monte Carlo simulations and the efficiency of the estimator is proven by comparing these results with the theoretical Cramér-Rao bound.