The allocation and scheduling of limited communication assets to an increasing number of satellites and other spacecraft remains a complex and challenging problem. Feasible schedules that address mission requirements over the short, medium and long term are inherently faced with the need to resolve conflicts when multiple missions contend for the same time windows of opportunity. This proposal is for an oversubscribed mission scheduler conflict resolution system that will rapidly provide feasible schedules using innovative boolean satisfiability methods and then leverage software agent personas to minimize or eliminate conflicts in scheduling. Agents will free up valuable human resources in the de-conflict loop, using an ask-offer process to negotiate future timeslots based on a deferred value function. Phase 1 will demonstrate the feasibility of applying mature SAT solver process to the rapid generation of feasible schedules and also demonstrate the potential for the application of semi-autonomous belief-desire-intent (BDI) agents to conflict resolution and user interface support. Phase 2 will evolve proof-of-concept prototypes into fully functional systems.
More »