A delay-tolerant network approach to satellite pickup and delivery scheduling

Abstract

An approach to job scheduling and task allocation in delay- and disruption-tolerant satellite networks, called Contact Graph Scheduling (CGS) is introduced. This method aims to minimize latency between the arrival of requests for satellite data products, and the delivery of those products, identifying efficient pickup and delivery paths through a resource- constrained time-varying network. Requests for goods are processed as soon as they arrive, resulting in the discovery of task distribution and product delivery routes, while being aware of both resource and operational constraints. CGS sits at the interface between Contact Graph Routing (CGR), the dynamic pickup and delivery problem (dPDP) and satellite scheduling. CGS provides a combined scheduling-routing solution, such that the route to be taken by data toward their destination informs the allocation of pickup tasks to network nodes. Results show that CGS can increase the number of fulfilled customer requests, while maintaining acceptable levels of delivery latency, compared to other, less burdensome methods. This study highlights the impact of contact reliability on CGS performance, suggesting that as uncertainty increases, so does the benefit of employing locally, rather than centrally, defined data routing strategies.