Tracking and data fusion in bi- and multistatic radar: a simulation study of geometry dependent uncertainties, filter selection and their influence

In bi-/multistatic radar, geometry dependencies are of common occurrence. This is also observed for measurement uncertainties. This paper discusses examples of bi-/multistatic radar systems, their geometry dependencies with respect to measurements and their input to a tracking and fusion system. For one of the systems, four simulated flight paths were used to evaluate the consequences of using different approximations to the measurement uncertainty. For all measures that were recorded from the Monte Carlo simulations, a closer approximation to the measurement uncertainty distribution resulted in better overall performance. Extended Kalman filters (EKFs) were used in single and two model filter versions as well as a two-model interacting multiple model (IMM) filter. For a simulated track consisting of straight-line segments and maneuvers, the IMM resulted in less average deviation values from true target values in the measured offsets in distance, velocity and heading.