Sensor Management using Expected Risk Reduction approach

This paper introduces an Expected Risk Reduction approach to Sensor Management and Multi-Target Tracking in a surveillance context that implements an IR-Radar sensor suite. Due to operational restrictions (for instance, electromagnetic emission constraints), it is assumed that there are more targets than given sensors are capable of tracking simultaneously when a radar emission control is applied. It is also presumed that an incorrect target classification entails a cost that is different for each target class. The Expected Risk Reduction is then applied to a simulated IR-Radar sensor management to preserve an acceptable level of kinematic accuracy on targets of high cost. Finally, empirical statistical tests show that a track on high priority targets is maintained better when the aforementioned approach is introduced than in the case of other conventional methods, such as the information gain approach or the round robin assignment.