Random Walk-Driven Encounter Routing for Enhanced Delivery Efficiency in DTNs

Abstract

Traditional networks face challenges in delivering messages when no direct path exists between the nodes. Delay Tolerant Networks address this issue through specialised algorithms, many of which leverage social metrics to select optimal relay nodes. While these approaches improve message delivery, they often incur high overhead costs. This paper introduces the routing protocol for DTNs, which aims to balance efficiency and performance by using three social metrics: Degree Centrality, Social Activeness, and Random Walk Encounter. Degree Centrality and Social Activeness both are already used in literature but we introduce a new social attribute called as Random Walk Encounter. The philosophy behind this protocol is to prioritise nodes that exhibit higher social connectivity and activity, ensuring that messages are forwarded through nodes with the best potential to enhance delivery performance while minimising the overhead associated with more traditional methods. The message is forwarded to the node with the highest Degree Centrality, Social Activeness, and Random Walk Encounter value, and this process continues until the message reaches its destination. The design philosophy behind this approach is to use social metrics that correlate with real-world human behaviours, optimising the selection of relay nodes for efficient data forwarding. We run the simulations with all three social attributes individually with same set parameters and simulation results show that proposed social attribute named as Random Walk Encounter outperforms Degree Centrality and Social Activeness, improving the Packet Delivery Ratio, with fewer hops and reduced overhead, at the expense of increased average latency.