Performance Bounds for Target Tracking: Computationally Efficient Formulations and Associated Applications
Marcel L. Hernandez · 2014
This chapter presents a review of recent developments in the calculation of mean square error (MSE) tracker performance bounds. It provides examples that demonstrate how such bounds can be used as the basis for performing online sensor management. The chapter deals with two applications in which controlling the posterior Cramér—rao lower bound (PCRLB) formulations provides an effective mechanism for online sensor management. The first application is concerned with the management of stationary sensors in order to track a single, potentially manoeuvring target. The second application is UAV trajectory planning, again in tracking a single target. In both applications, the efficient formulation of the PCRLB, together with other efficiency savings, reduce, the computational overhead of the two algorithms to a minimum. Finally, the chapter points out that the recent efficient formulations of the PCRLB offer a powerful technique for performing sensor management in target tracking, whilst allowing for real-world complexities.