Use of store and forward metrics for service-oriented routing in satellite constellations

Routing policies in traditional networks are based on interactive mechanisms that consider a permanent simultaneous communication between the source and the destination nodes. In addition, the metrics associated to these networks (throughput, end-to-end delay, jitter, etc.) are based on this assumption. Systems deployed in challenging environments are faced to intermittent connectivity as a consequence of environmental conditions, mobility or equipment failure. The delay tolerant network (DTN) architecture [1], [2], based on store and forward (S&F) mechanisms is designed to accommodate these conditions. As traditional metrics can not be used for characterizing routes in this kind of networks, a framework of new S&F metrics has been proposed in [3]. In this paper, we present an algorithm to compute routes in a deterministic S&F system made of a low earth orbit (LEO) satellite constellation. Three different routing policies are applied depending on the offered services. The results show the interest of S&F metrics for service-oriented routing policies.