3D Pseudolinear Target Motion Analysis From Angle Measurements

The paper presents a new pseudolinear estimator for 3D target motion analysis by a single moving ownship collecting azimuth and elevation angle measurements. The 3D pseudolinear estimator is derived as a small-noise approximation to the maximum-likelihood estimator and consists of a 2D pseudolinear estimator for the xy-components of the target motion parameters and a least-squares estimator for the z-component. To improve the poor bias performance of the 3D pseudolinear estimator, alternative estimators are proposed employing bias compensation and weighted instrumental variables. A selective-angle-measurement implementation of weighted instrumental variables is presented to maintain a strong correlation between the instrumental variable matrix and the data matrix in the presence of large measurement noise. The performance advantages of the selective-angle-measurement weighted instrumental variable estimator over the conventional maximum likelihood estimator are demonstrated via simulation examples.