Optimal Tracking of Stationary Target by Moving Platform with Doppler Information Only in 3D
Liat Peled-Eitan, Ilan Rusnak · 2021
The problem of estimating the three-dimensional position of a stationary target given Doppler frequency measurements from a moving platform only is addressed. The existing approach for estimation a stationary target location from Doppler only information is by using a multi-sensor network. The motion of the sensor makes the problem observable without the need for multi-sensors. This is as the motion of the sensor makes a full continuum of sensors. The measurement of the closing velocity, proportional to the Doppler frequency, is a nonlinear function of the Cartesian location of the target. The recently introduced Optimal Nonlinear Recursive Least Square (RNLS) filter, the State-Dependent Differential/Difference Riccati Equation (SDDRE) based filter, and the Extended Kalman Filter are compared on a common basis, for the Doppler measurement only position estimation. This work shows that despite the possibility of multiple solutions and the dependence of the estimation quality on the sensor's trajectory, optimal tracking of a stationary target from Doppler measurements only by a single sensor is feasible. The superior performance of the optimal Recursive Nonlinear Least Square filter is demonstrated by simulations.