Message dropping policy in congested social Delay Tolerant Networks

Abstract

Congestion in DTNs happens when the forwarding path to destination does not exist and a node runs out of buffer space. Several works have shown that the situation is hard to mitigate. Consequently, to choose the right message to drop is a crucial task. Moreover, there has not been much attention to the buffer management policy in the Social-DTNs contexts. Therefore, an efficient buffer management policy for Social-DTNs is needed to achieve a better performance for communication in challenging disruptive scenarios. In this paper, we propose a novel message dropping policy for Social-DTNs. Our proposed policy utilizes social relation of nodes to avoid dropping meaningful messages in congested social-DTNs. We have implemented and evaluated our policy on the ONE simulator, and used real contact traces obtained during a conference. Our evaluation results show that our proposed policy can significantly reduce the number of overhead and can outperform the traditional policy under investigated scenarios.