The Research Of Data Distribution Protocol In Urban VANETs

The Vehicular ad hoc networks(VANETs)has attracted increasing attentions due to their potential to improve driving safety and traffic efficiency.In VANETs,due to the high mobility of vehicles and the impracticability of maintaining the entire network topology information,the traditional data distribution protocols cannot meet the basic requirements for data distribution.Therefore,how to improve the performance of network communications in VANETs is a research hotspot in the academic community.For the characteristics of the highly dynamic and unreliability in VANETs,this thesis not only studied broadcast protocols based on emergency data,but also considered unicast protocols for non-emergency data.The contributions of this thesis are summarized as follows:In non-emergency data distribution strategies,intersection-based data distribution strategies are more adaptable to the characteristics of VANETs.The thesis considers the dynamics of the topology of urban environment networks and proposes a connectivity routing protocol based on transmission cost.In the free-connected state,the connectivity probability of a road segment is deduced theoretically by analyzing the data packet transmission process in the lane.Taking into account the location of the road section will have an important impact on the routing performance,the thesis proposes the concept of road segment correlation and design a new routing metric for the road section correlation,which analysis transmission overhead from the source node to the destination node.At the intersection,this route metric can help the vehicle dynamically select the street sequence.After selecting the optimal path,the position prediction method is used to predict the position information of neighboring nodes,which effectively reduces the possibility of forwarding packets to invalid neighboring nodes and improves the delivery rate of data packets.The numerical simulation results show that our proposed protocol has better performance in data packet transmission rate and endto-end average delay.Due to the high mobility of the vehicle and the complex communication environment,it is easy to select an inappropriate next hop,resulting in a worsening result.To deal with this issue,in this thesis,a probabilistic broadcasting for emergent message dissemination is proposed.This article defines the concept of Region of Sensitivity(Ro S).This is to prevent the infinite diffusion of messages and save network overhead.This thesis proposes the expectation relay node,which is used as the selection guide of the relay node in the broadcast process.It is the expected relay node(ERN)that considers three key factors: distance,link availability,packet reception probability,and the sending end assigns weight to each candidate node.Based on the weight probability metric,the sender selects the node with the highest weight probability as the desired relay node.When it is expected that the relay node cannot successfully receive or forward the data packet,other sub-optimal nodes will assist in forwarding the data packet.The numerical simulation results show that our proposed protocol can greatly improve the reliability of emergency information dissemination.The data distribution protocols proposed in this thesis can effectively distribute data according to different application requirements.The data distribution protocols can save network resources,provide high-quality services and have a good application prospect.