Delay Tolerant Networks (DTNs) are novel wireless mobile networks, which suffer from frequent disruption, high latency, and the lack of a complete path from source to destination. Vehicular Delay Tolerant Network (VDTN) is a special type of DTNs with vehicles as nodes. In VDTN, most nodes have specific movement patterns, however, traditional routing algorithms in DTNs do not take this characteristic into considerations very well. In this paper, a new routing algorithm based on Bayesian Network (BN) is proposed to construct the prediction model, which intends to predict the movement patterns of nodes in the real VDTN scenarios. Firstly, a comprehensive BN model is established, where more attributes of nodes are selected to improve the accuracy of the model prediction. Then, considering the complexity of the structure learning problem of BN, a novel structure learning algorithm, K2 algorithm based on Genetic Algorithm (K2-GA), is proposed to search the optimal BN structure efficiently. At last, Junction Tree Algorithm (JTA) is adopted in the inference of BN, which can accelerate the inference process through variable elimination and calculation sharing for large scale BN. The simulation results show that the proposed VDTN routing algorithm based on the BN model can improve the delivery ratio with a minor forwarding overhead. INDEX TERMS Vehicular delay tolerant network, routing algorithm, Bayesian network, genetic algorithm, optimization.

Delay-tolerant networks (DTNs) are wireless mobile networks, which suffer from frequent disruption, high latency, and lack of a complete path from source to destination. The intermittent connectivity in DTNs makes it difficult to efficiently deliver messages. Research results have shown that the routing protocol based on reinforcement learning can achieve a reasonable balance between routing performance and cost. However, due to the complexity, dynamics, and uncertainty of the characteristics of nodes in DTNs, providing a reliable multihop routing in DTNs is still a particular challenge. In this paper, we propose a Fuzzy-logic-based Double Q -Learning Routing (FDQLR) protocol that can learn the optimal route by combining fuzzy logic with the Double Q -Learning algorithm. In this protocol, a fuzzy dynamic reward mechanism is proposed, and it uses fuzzy logic to comprehensively evaluate the characteristics of nodes including node activity, contact interval, and movement speed. Furthermore, a hot zone drop mechanism and a drop mechanism are proposed, which can improve the efficiency of message forwarding and buffer management of the node. The simulation results show that the fuzzy logic can improve the performance of the FDQLR protocol in terms of delivery ratio, delivery delay, and overhead. In particular, compared with other related routing protocols of DTNs, the FDQLR protocol can achieve the highest delivery ratio and the lowest overhead.