Clustering routing method based on social characteristic in campus DTN

Campus networks are featured by intermittent connection and dynamically changing topology. Due to these factors, a campus network becomes a typical scenario of delay-tolerant network (DTN). In order to ensure routing efficiency in a campus DTN, in this paper, we record the movement paths and mutual encounter numbers of the nodes and regard the quantified social characteristic as the weight of the edges between node pairs, and then a weighted graph of the network topology is obtained. Clusters including nodes with high intimacy between each other are extracted utilizing a topological cutting method. Subsequently, messages are forwarded according to the nodes' degrees of cohesion and outreach. As a result, we propose a clustering routing method based on social characteristic in DTN. Simulations based on the synthetic mobility pattern and real trace are conducted in ONE. The results show that the method can achieve about 10% higher delivery ratio and 200s less average delay than the other routing protocols. In other words, the method improves the delivery ratio and reduces the average delay in DTN.