Adaptive infection recovery schemes for multicast delay tolerant networks

Conventional Delay Tolerant Networks (DTN) use "store-carry-forward" paradigm to pass the message between the nodes that meet occasionally which results in intermittent connectivity. Once the message meets the destination, the network initiates the so called "infection recovery process" in order to remove the delivered messages from the rest of the nodes. This process starts as soon as the message reaches the first destination which, in the case of multicast sessions, may reduce the chances that the other destination nodes receive the message. In this paper, we present an analytical framework to study the performance of different infection recovery schemes for multicast DTN. New adaptive recovery schemes are developed where the recovery probability is adjusted to the multicast traffic. The performance of these new algorithms is compared to a number of unicast recovery schemes modified for multicast DTN, which also represents a contribution of this paper. Our analytical framework can be easily extended to model the recovery process for different multicast routing schemes. Numerical results show that by adaptive immune, immune-TX and vaccine schemes the delivery delay can be reduced up to 75% compared to the conventional schemes. By timeout recovery scheme, when applied to multicast session, the reduction in the delivery delay can reach up to 90% at the expense of larger recovery delay.