Opportunistic message forwarding in self organized cluster based DTN

Delay Tolerant Networks (DTNs) is that class of challenged network that works on existing infrastructure. Any two nodes that move within each other's transmission range can exchange messages. This may form the basis for future forwarding based on past contact history. Based on these identifiable patterns one may predict contacts in near future. Some existing schemes use utility functions to steer the routing in the right direction but are limited, and is unable to capture enough information of the network. In this paper, the Trust Model is further extended to identify cluster heads in DTN environment that results in the optimal utilization of the node's resources. Due to high network mobility, it is assumed that the node's contacts are unpredictable so that it can reflect the most efficient way of message movement. Further, the identification of the cluster head enhances delivery ratio and reduces delivery latency. Fewer messages are aborted in case of proposed algorithm as compared to FirstContact. There is 7.7% reduction in the number of messages in the network as compared to FirstContact routing protocol. The proposed routing protocol is simulated using real-world mobility traces. Also, the minimum hop count is observed that leads to lesser overhead. The proposed work demonstrates the applicability of the trust to compute and forward the message to DTN's using cluster head. This implicitly applies security to an acceptable level. Experimental work shows research direction towards resource optimization in DTN network.