Multistatic pseudolinear target motion analysis using hybrid measurements

This paper presents a new hybrid pseudolinear estimator (PLE) for target motion analysis of a constant velocity target in the two-dimensional plane using angle-of-arrival, time-difference-of-arrival and frequency-difference-of-arrival measurements obtained from spatially distributed stationary passive receivers. The hybrid PLE is developed by linearizing the nonlinear measurement equations in the unknown target motion parameters. The resulting estimator is not only closed-form and has low computational complexity, but is also free from nuisance parameters, therefore avoiding the problems arising from the dependence of the nuisance parameters on the target motion parameters. However, the noise injected into the PLE data matrix causes biased estimates. To address this, a bias-compensated PLE is proposed based on an asymptotic bias analysis of the hybrid PLE. This estimator is then incorporated into a weighted instrumental variable (WIV) estimator to obtain asymptotically unbiased estimates of the target motion parameters. The WIV estimator is shown to be asymptotically efficient both analytically and through numerical simulation examples. Furthermore, it is observed that the WIV estimator performs similar to the computationally demanding maximum likelihood estimator, closely achieving the Cramer-Rao lower bound and producing negligible bias at moderate noise levels. (C) 2016 Elsevier B.V. All rights reserved.