The goal of the proposed effort is to develop a fully adaptive radar resource allocation framework for concurrent detection and tracking of targets. While the methods we will develop are general, we will specialize our analysis to ground targets by including clutter and using kinematic models relevant to ground targets. In Phase I, we will develop both task-driven and information-driven metrics in a single radar sensor setting and compare their performance both in simulation and in a baseline laboratory experiment. We will develop a benchmark simulation testbed to enable comparison of the algorithms against each other and against standard techniques. We will identify the relevant tuning parameters, computational requirements, and the merits and deficiencies of each objective function approach. Particular attention will be given to the dependence of the optimization with respect to the scale of the problem in terms of the number of sensors and targets. In Phase II, we will extend the approach to include simulations and experiments involving a suite of distributed radars and multiple moving targets.
