Computing maximal track clusters for sensor resource management

In this paper, we present an approach to clustering tracks into maximal groups for sensor resource management. The motivation is that sensor tasks include ones for revisiting existing tracks, and one would like to cluster these tracks into groups that are simultaneously observable. The sensor resource manager can then schedule the groups for observation. Note that the feasible groupings are constrained by the observable region of the sensor. We illustrate our approach for two different types of sensors with different observable regions: mechanically-steered sensors such as gimbaled electro-optical or dish radars and electronically-steered radar arrays. Our numerical results demonstrate that our approach to clustering tracks into maximal groups reduces the number of tracks groups that need to be observed as compared to more conventional means for clustering tracks.