An optimal packet scheduling and load balancing algorithm for LEO/MEO satellite networks

LEO/MEO constellations of communication satellites have recently been proposed as a powerful tool for improving internet performance and for extending it beyond the earth (InterPlanetary Internet). Several aspects of such usage must still be investigated. Among them, proper routing and MAC protocols play a prominent role. In this paper,we assume that the MAC protocol is an MF/TDMA one (as usual in satellite communication), and we consider the problem of assigning packets to a set of shortest paths in a satellite constellation, with the goal of minimizing the overall packet scheduling problem. We present an optimal polynomial time algorithm which works offline and balances the load over all possible shortest paths, allowing a minimum schedule length for the entire constellation to be found in polynomial time.