An evolutionary approach towards contact plan design for disruption-tolerant satellite networks

Abstract

Delay and disruption tolerant networks (DTNs) are becoming an appealing solution for satellite networks where nodes can temporarily store and carry in-transit data until a link with a suitable next-hop becomes available. Since satellite trajectories and orientation can be predicted, on-board routing schemes can base these forwarding decisions on a contact plan comprising all forthcoming communication opportunities. In general, contact plans are previously calculated on ground where their design can be optimized to consider not only available spacecraft resources but also the expected traffic which is largely foreseeable in space applications. Despite optimal contact plan design procedures exist, their computation complexity might result prohibitive even for medium-sized satellite networks. In this work, we propose an evolutionary algorithm to provide sub-optimal yet efficient and implementable contact plans in bounded time. In particular, we depict specific strategies such as encoding and repairing techniques to later evaluate the algorithm performance in a typical scenario demonstrating its usefulness for planning future DTN-based satellite networks.