Research On Multi-attribute Decision-making VDTN Routing Protocol Based On Road Clustering

In recent years,the number of urban vehicles has been increasing rapidly,which has caused the frequent occurrence of road congestion and traffic accidents.Under this background,Vehicular Ad hoc Networks(VANETs)has emerged,it is a significant component of intelligent transportation system for providing various in-car services for users.Vehicular Delay Tolerant Network(VDTN)can enable the stable transmission of messages in the VANETs of intermittent connection,and thus has received the attention by more and more researchers.Most existing VDTN routing algorithms ignore controlling the redundant messages in the local network,which may worsen the routing performance by distributing redundant message copies in a local network far away from the destination.A Multi-Attribute Decision-making Algorithm Based on Clustering(MDBC)is proposed to control redundant message copies in local network.Then,with the proposed routing algorithm,a cooperative approach based on the query-reply mode for traffic condition estimation is proposed to further expanding the scope of the vehicle perception and obtain the traffic condition of more road through cooperative vehicle-to-vehicle communications.Meanwhile,in order to evaluate the accuracy of the proposed approach accurately,a Vehicle Mobility Model Based on the Spatial and Temporal Distribution of the Real Traces(STRT)is proposed.The main research results are as follows:(1)A Multi-Attribute Decision-making Algorithm Based on Clustering is proposed.Firstly,a node clustering method based on roads is proposed.Then,MDBC enables nodes aware of local message distribution which is used as a basis for forwarding messages to control redundant message copies in local network.Meanwhile,MDBC establishes a multi-attribute decision-making model using the neighbor node attribute,the own attribute and the road attribute in order to make the routing decision more accurate.(2)Based on the proposed routing algorithm,a cooperative approach based on the query-reply mode for traffic condition estimation is proposed.Firstly,each vehicle evaluates the road traffic in the passing region based on wireless location technology of moving target and by exchanging information with neighbor vehicles so as to maintain a global network traffic state tables.Then,vehicles can obtain and predict the traffic state of the interest road segment.(3)In order to reflect the driving behavior of vehicles and the traffic flow patterns in the real scene,a Vehicle Mobility Model Based on the Spatial and Temporal Distribution of the Real Traces is proposed.Firstly,a quadratic function model is put forward to simulate the correlation between the traffic flow and the time on the basis of real vehicle trajectory data mining in Shanghai early peak traffic flow with the time.Then,a route planning strategy based on region popularity is proposed.Last,the speed of the vehicle is updated periodically based on Gauss-Markov model during driving.(4)This paper implements the proposed routing algorithm,traffic state sensing method and vehicle mobility model in the Opportunistic Network Environment(ONE)simulator.Through experiments,the feasibility and efficiency of the proposed schemes are verified compared with existing technology.