Optimal estimation of false measurement density

The task of tracking targets using information from sensors can be very difficult. In addition to measurements from the required target, sensors produce measurements which result from random noise, clutter, countermeasures and interference. Such measurements are termed ´false´. As a result, it is not possible to distinguish with certainty the origin of sensor measurements, and ideally tracking algorithms need to take this uncertainty into account. One way of accommodating false measurements is to construct a model for their behaviour and estimate any parameters which appear in that model. A statistical model for false measurements is described. A Bayesian technique is defined to derive an optimal false measurement density estimator. This estimator is simple to update with new measurement data and allows for the degradation of knowledge between measurements.