Congestion modeling and management techniques for predictable disruption tolerant networks

FRAIRE, JUAN; MADOERY, PABLO; FINOCHIETTO, JORGE; BIRRANE, EDWARD

Clearwater Beach, Florida, USA

Conferencia; The 40th IEEE Conference on Local Computer Networks (LCN); 2015

IEEE

Delay and disruption tolerant networks (DTNs) arebecoming an appealing solution for extending Internet boundariesso as to embrace disruptive communications. In particular,if node trajectory can be predicted as in space networks,routing schemes can take advantage of the a-priori knowledgeof a contact plan. Despite mechanisms such as Contact GraphRouting (CGR) exist, they might derive in harmful overbookingof forthcoming contacts, also known as congestion. In order totackle congestion, we initially formulate the problem by meansof a linear programming (LP) model so as to establish anupper theoretical bound of performance. Next, we survey existingcongestion mitigation mechanisms for predictable DTNs to latercontribute with a CGR extension named PA-CGR. Finally, inthe pursuance of an optimal congestion avoidance approach, wealso propose and evaluate in a realistic scenario a novel multigraphtechnique (MG-CGR) that outperforms existing solutionsby exploiting traffic predictability.