Congreso: 2018 IEEE Radar Conference (RadarConf18), April 23-27, 2018, Oklahoma, USA.    

Autores: Luis Úbeda Medina, Ángel F. García Fernández, Jesús Grajal de la Fuente.

The fully adaptive radar framework aims to use the available information from the scenario in which a system
is deployed to adaptively change its configuration with the intention of achieving a performance improvement. The use of
this strategy relies on an optimization procedure, usually based on the minimization of the predicted conditional Cramér-Rao lower bound, which in some difficult scenarios can result in misperformance of the system due to a lack of robustness. In this paper, we present a novel approach based on a linear-Gaussian approximation, to carry out the optimization procedure inherent in the fully adaptive radar framework that successfully avoids these robustness issues. This method can be easily implemented using sigma-point integration methods. We demonstrate the performance of the proposed approach through simulations in a single target tracking scenario using a sensor network.