A batch processing algorithm for moving surface target tracking

This paper develops a batch processing algorithm that can be used to track a constant velocity surface target. The purpose of this algorithm is to facilitate passive tracking when sensor-target geometry is poor, which can prevent the convergence of a recursive estimator. The target's position is considered to be the output of an ordinary differential equation having unknown parameters to be estimated. This contrasts with the model used for the design of recursive estimators such as a Kalman filter where the target's position is the output of a dynamic system driven by white noise. Batch processing of all sensor measurements and Iterated Least-Squares (ILS) are used to estimate the target model parameters. Numerical integration is used to propagate the target's position and the Jacobian needed by ILS. Simulation results are shown for a maritime surveillance mission.