Multi-Copy Routing with Trajectory Prediction in Social Delay-Tolerant Networks

Routing in Delay-Tolerant Networks (DTNs) remains a challenging problem due to sporadic connectivity and high delays. To deal with this issue, researchers have investigated multi-copy schemes with predicting future contacts. However, most of the previous work has focused on the prediction of future contacts, without sufficiently considering contact times. This paper proposes McRTP, a multi- copy routing protocol with trajectory prediction for social DTNs. Based on estimating the probability distribution of future contact times, McRTP evaluates and selects multiple paths with the highest delivery probability for routing. Also, to control network traffic overhead, we develop a copy count assignment scheme limiting the number of message copies in networks. A simulation study shows that, with the estimation of both contact probability and contact times, and the path selection scheme, our McRTP outperforms traditional DTN multi-copy routing schemes, especially in sparse networks.