Autonomous spacecraft navigation using distributed, locally optimal filters

The cost and vulnerability of ground based spacecraft tracking systems and limitations on the weight and cost of the sensor complement on an individual spacecraft combine to encourage the use of range measurements between spacecraft for autonomous spacecraft navigation. Optimal use of such measurements, however, may require excessive computer time and memory. An autonomous navigation algorithm employing distributed, locally optimal filters to process range measurements has been evaluated by means of covariance analysis. Results show that instability due to multiple feedback can be eliminated by altering the locally perceived sensor variance. If a hierarchical structure is chosen, the efficacy of a rank assignment scheme is strongly dependent on relative orbital position.