Sociality-Aided New Adaptive Infection Recovery Schemes for Multicast DTNs

Abstract

Delay-tolerant networks (DTNs) consist of nodes moving around and occasionally coming into each other's proximity. During the limited proximity time, nodes can exchange data; this can result in a very slow data dissemination process that is usually governed by a replication-based mechanism. However, due to the long propagation delay and the large overhead associated with the replication approach, DTN delivery performance is neither efficient nor effective. To reduce the overhead, which becomes a critical aspect particularly when addressing a multicast scenario, infection recovery mechanisms have been proposed to control and reduce the number of packet copies circulating through the network. This, however, has the cost of decreasing the chances of delivering packets to all destinations. In this paper, adaptivity in infection recovery is addressed. This represents a viable solution to make transmission more reliable, hence delaying the activation of the infection recovery procedure, depending on the number of nodes, destinations, and the time. Moreover, we also propose to exploit an additional feature in data multicasting, i.e., socially aided data dissemination, where the packet dissemination procedure is not trivially epidemic, but rather exploits the intrinsic sociality of users and their interests to reduce the delivery overhead and speed up the multicast process. More specifically, we consider a procedure where users are not regarded as individual members of the network but can be aggregated into groups sharing interests, and their sociality helps the data dissemination procedure. Results of our analysis show that new sociality-aided adaptive recovery schemes can speed up the delivery process.