Impact of persistent storage on the DTN routing performance

The store, carry, and forward paradigm of the Delay- Tolerant Network (DTN) architecture enables a node to carry messages for a long period of time. This long-term storage is supported by the DTN architecture with the usage of persistent storage; however to the best of our knowledge, the routing/scheduling framework that incorporates support for persistent storage has not been addressed much in the DTN literature. In this paper, we investigate the impact of persistent storage on the routing performance over different buffer scheduling policies. Our extensive simulation studies demonstrate that they exhibit an improvement in delivery ratio, but with a compromise on delivery delay. This shows the pressing need for a new scheduling policy to tap the complete potential of the persistent storage. To this end, we propose a Time in Primary Scheduling (TiPS) policy with two variants (one using local information and the other using global information) that outperforms the contemporary buffer scheduling policies with respect to the persistent storage framework. © 2012 Springer-Verlag.