Comparison of two single-step, myopic sensor management decision processes applied to Space Situational Awareness

This paper describes a sensor tasking, or sensor network management approach for a multiobject, multi-sensor tracking problem analogous to the monitoring of resident space objects known as Space Situational Awareness (SSA). In the SSA problem large discrepancies between satellites tracked and resources available to track them create difficulties in maintaining accurate satellite position and uncertainty estimates. Long periods of either inability to make observations (due to line-of-sight access) or unavailability of sensors (due to scheduling constraints) necessitates the need to intelligently determine which satellites should be observed and which should be ignored at various times, a process known as sensor tasking. To conduct this tasking, two information theory-based utility metrics are used, where one is a measure of absolute information gain and the other a quantification of relative information gain. These metrics are implemented in a single-step optimization problem in order to maximize total information gained over a series of observations from five sensors (four ground-based, one orbiting) measuring the range and azimuth of several satellites. Using a simple simulation of the estimation and tasking components of the SSA problem, these metrics are implemented in conjunction with an extended Kalman filter or unscented Kalman filter to obtain the satellites state and uncertainty estimates. Comparisons are made between the two methods of tasking and two estimators in order to determine which combinations of filters/tasking produce the most desirable tracking performance.