AAS 07-150 A GREEDY RANDOM ADAPTIVE SEARCH PROCEDURE FOR OPTIMAL SCHEDULING OF P 2 P SATELLITE REFUELING

All studies of peer-to-peer (henceforth abbreviated as P2P) satellite refueling so far assume that all active satellites return back to their original orbital positions after undergoing fuel exchanges with the passive satellites. In this paper, we remove this restriction on the active satellites, namely, we allow the active satellites to interchange orbital slots during the return trips. We formulate the problem as a three-index assignment problem in a tripartite constellation graph. We use a greedy random adaptive search procedure (GRASP) to determine the optimal P2P refueling schedule between the active and the passive satellites. It is shown that the proposed methodology leads to considerable fuel reduction over the baseline P2P refueling strategy.