Tracking Performance Self-Determination and Prediction for Sensor Data Fusion and Resource Management

We propose to develop a tracking performance self-determination methodology with which a tracker can assess its operational conditions and predict its performance. With the self-determined performance metrics, a sensor resource manager can efficiently reroute or reschedule the radar activities so as to resolve data association ambiguity and ensure the overall track quality. In Phase I, a model for a tracker integrated with a sensor resource manager will be established for closed-loop simulation. Self-determined performance metrics will be introduced in comparison to absolute performance metrics which require the truth trajectory not available in run time. In addition to analytic derivation for simple cases, complex performance curves will be obtained in designed experiments via computer simulation. Key influence factors will be assessed including target maneuver and data association. A deferred updating with an interval smoother is proposed to solve data association ambiguity of crossing targets. In Phase II, the Phase I-validated methods will be extended with a comprehensive set of metrics, information needs, and theoretical performance capabilities over various operating conditions.