A Statistical Analysis and Temporary Cluster Based Routing Algorithm for Delay Tolerant Networks

Delay Tolerant Networks are characterized by frequent network topology partition, limited resource, extremely high latency, etc. Consequently successful deliveries of messages in such networks face great challenges. In this paper, by capturing the temporary clusters, we timely use the temporary end-to-end paths to directly and successfully deliver messages. Besides, by collecting a large number of encounter history information as the samples, we use the methods of statistical analysis to objectively and accurately evaluate encountered nodes, thus selecting fewer but better relay nodes to spread messages. Finally based on the above schemes, a Statistical Analysis and Temporary Cluster based routing algorithm (SATC) is proposed to improve the routing performance. Extensive simulations have been conducted and the results show that SATC can achieve a higher delivery ratio and a fewer hop count compared to Epidemic and PRoPHET. Furthermore, its overhead ratio is 70% and 65% less than Epidemic and PRoPHET respectively.