Managing Routing Scalability in Space DTNs

Routing in the Space Internet: A contact graph routing tutorial

“…In 2016, Burleigh et al proposed an opportunistic enhancement to CGR so that unplanned contacts could be correctly reacted upon and included in the routing decisions [82]. Dr. Ruhai Wang et al then presented the first investigations into CGR scalability [83], which motivated later contributions by Madoery et al via efficient forwarding [84] and region-based approaches [85], [86]. Initial CGR reliability studies followed in 2017.…”

“…While software optimizations might mitigate this problem somewhat, adaptations that would enable CGR to scale up to larger networks over extended time-horizons should be explored. Hierarchical inter-regional routing: Hierarchical interregional routing (HIRR), proposed in [85], [86], aims at reducing the scope of route computation by limiting the topology knowledge provided to each DTN node in the network -in other words, reducing the contact plan size. In order to control the routing computation effort, the network can be divided into multiple regions as illustrated in Fig.…”

“…A bundle that is destined for another node in the source node's own region is routed to that node by CGR; a bundle that is destined for a node is another region is routed via CGR to the passageway node that is a member of the destination region (and thence to the destination node by CGR) -or is a member of a region from which the bundle can ultimately be routed to the destination region by these same mechanisms. An initial inter-region routing algorithm was proposed in [85], and an auto-discovery approach was described in [86]. Further evaluation, potential optimization, and efficient region design procedures motivate further research efforts in this important domain.…”

Routing Heterogeneous Traffic in Delay-Tolerant Satellite Networks

“…In addition to this, other improvements include congestion mitigation techniques [10], route table management strategies and the incorporation of Yen's algorithm [11]. Furthermore, the incorporation of opportunistic [12] and probabilistic contacts [13], the design of new schemes to enhance the scalability [14], a spanningtree formulation to compute routes to several destinations [15], and a partial queue information sharing [16], as well as many other contributions.…”