Command and Control of a Cluster of Satellites

There is an increasing desire in many organizations, including NASA and the DoD, to use constellations or fleets of autonomous spacecraft working together to accomplish complex mission objectives. At the Air Force Research Laboratory’s (AFRL) Space Vehicles Directorate we are investigating and developing architectures for commanding and controlling a cluster of cooperating satellites. For many space missions, large monolithic satellites are required to meet mission requirements. In many cases this results in costly satellites which are more complex, more susceptible to failure, and which have performance characteristics that are less than optimal due to realistic physical size limitations. Recently various organizations have begun to explore how distributed clusters of cooperating satellites can replace their larger monolithic counterparts resulting in an overall cost reduction, enhanced mission performance, and increased system fault tolerance. Large clusters of satellites flying in formation are required to have some level of on-board autonomy in order to: fly within specified tolerance levels; perform collision avoidance; address fault detection, isolation, and resolution (FDIR); share knowledge; and plan and schedule activities. In addition, from an operations standpoint, commanding and controlling a large cluster of satellites can be very burdensome for ground operators. At AFRL we are addressing these issues by development of an on-board Cluster Management system which will, in essence, provide the capability to treat a cluster of satellites as a single virtual satellite. A systems level approach is being taken, therefore from a ground perspective the ground control station must also be able to treat the cluster as a virtual satellite. [1] This paper will describe our Cluster Management system, which is the intelligent entity that is responsible for making cluster level decisions. It is this piece of software that enables a satellite cluster to function as a virtual satellite. The cluster manager functionality can be broken down into the following five areas: