An Improved Probabilistic Routing Algorithm Based on Moving Direction Prediction in DTNs

Abstract

In a delay-tolerant network (DTN), nodes use a store-and-forward mechanism to forward messages from a source node to an encounter node until the message reaches the destination node. The traditional probabilistic routing protocol using history of encounters and transitivity (Prophet) is one of the important routes in DTN, to decide whether the node carrying the message should pass the message to the encountering node. The scene of this paper focuses on some DTNs with a large moving area. Most of the default communication devices are mobile, and the moving direction is random. If the encountering node is close to the destination node, the next encounter is still close. When trending, we will record the movement direction between nodes and use the prediction of the movement direction to make more reasonable decisions on the routing algorithm. Therefore, an improved probabilistic routing algorithm (M-Prophet) based on moving direction prediction is proposed, which is based on the node transmission probability update stage of the traditional Prophet algorithm and improves by updating the encounter probability according to the moving direction between nodes. M-Prophet first records the delay of the encounter between nodes and defines the moving direction according to the comparison between the last delay and the delay of this encounter, so as to design a new node delivery probability formula. The final simulation results show that the M-Prophet algorithm proposed in this paper can improve the performance of the message delivery success rate by up to 11% compared with the traditional Prophet algorithm and reduce the average delay by about 10 s.