Multisensor bearings-only tracking using the MDL criterion

The problem of multitarget tracking using bearings-only measurements is addressed, when the number of targets is unknown a-priori. The minimum description length (MDL) criterion of Rissanen (1983) is first chosen as a natural way to determine the number of targets when a prior distribution is unavailable. However, it is shown that the MDL criterion lends to overestimate the number of targets, and hence a modified criterion is proposed. The resulting algorithm corresponds to the computation of joint maximum likelihood estimates of target states and associations, with an additional penalty term to prevent overparameterization. The problem of data association is solved using a set of parallel simulated annealing algorithms over the sensors and scans. As the associations are formed by annealing, a conventional nonlinear programming algorithm simultaneously estimates the target states (position and velocity). The consistency of the new estimation criterion is proven analytically in the case of a clean environment. Simulation results are presented which compare the tracking performance of the MDL and modified estimation algorithms, for cases with and without clutter