Research On Routing Algorithms Of Vehicular Delay Tolerant Networks Based On Bayesian Network

Delay Tolerant Network(DTN)is a new type of wireless sensor network.Compared to traditional stable networks based on TCP/IP,DTN typically faces issues such as frequent interruptions,high latency,and unstable connections between nodes.As a special application of DTN,Vehicle Delay Tolerant Network(VDTN)is a special DTN with vehicles as nodes.Considering that the movement of vehicles in VDTN is influenced by human consciousness,different types of vehicles may have different motion ranges and modes.And from a time perspective,the motion of the vehicle exhibits significant dynamism.However,traditional DTN routing algorithms have not taken into account the characteristics of VDTN,so researchers need to propose more effective routing algorithms based on the characteristics of VDTN.In this thesis,firstly,the relevant research background,theoretical knowledge and key technologies of VDTN,especially the current research status of VDTN routing algorithms are analyzed.And then the concept and some basic knowledge of Bayesian network(BN)involved in this paper,which provide a basis for the next design of VDTN routing algorithm based on Bayesian Network,are introduced.Aiming at the unique operating modes and network characteristics of vehicles in VDTN,a VDTN routing algorithm based on clustering Bayesian network(BN)is proposed by taking advantage of the characteristics of spatial correlation between vehicle trajectories.Firstly,a definition of distance between trajectories is given based on Hausdorff distance according to the space characteristics between trajectories.Secondly,a K-means algorithm is used to cluster the historical data sets of vehicle trajectories(KTC algorithm),so that the vehicle trajectories with spatial-temporal similarity can be grouped into one class;Thirdly,a K2 algorithm is used to build different BN models for different classified vehicle trajectories historical data sets to improve the accuracy of BN models;Finally,the message delivery level of the vehicle node inferred from the BN model is used as the basis for message forwarding.The simulation results show that the proposed VDTN routing algorithm based on the clustering BN can effectively improve the message delivery ratio and reduce the delivery delay.Further considering the spatial characteristics and dynamics of vehicle movement modes in VDTN,a VDTN routing algorithm based on Dynamic Bayesian Network is proposed.On the basis of the VDTN routing algorithm based on clustering Bayesian Network,the Dynamic Bayesian Network(DBN)model is used to realize the consideration of time latitude,which can consider the impact of time dynamic factors in more detail,making routing decisions more accurate.K2~+algorithm and KTC based DBN structure learning algorithm(KDBN)are proposed on the basis of K2 algorithm for the structure learning of DBN.The simulation experiment shows that the performance of the proposed routing algorithm has been significantly improved.The research results of this thesis can provide ideas for the research of routing algorithms in VDTN,and have good theoretical value and application scenarios.