VANET Routing Protocol Based On Urban Three-Dimensional Traffic Scenarios

The widespread use of vehicles provides convenience to people and creates social problems such as traffic congestion and accidents.The Intelligent Traffic System(ITS)is an internationally recognized solution to urban traffic problems.It plays an active role in decreasing urban traffic pressure,improving travel efficiency,and ensuring traffic safety.The large number of cars makes urban space resources very limited.Since the traditional transportation structure can not match the current urban transportation needs,more and more cities are focusing on urban three-dimensional transportation.The complicated road layout in3 D sceneries poses new issues for ITS when it comes to establishing dependable routing.Firstly,the research background and status of Vehicular Ad Hoc Network(VANET)are introduced in this thesis.The design principles,metrics,and classification criteria of routing protocols are investigated.And then,the classical routing protocols in different scenarios are investigated.The characteristics of the Three-Dimensional Vehicular Ad-hoc Network(3DVANET)and its routing protocols are discussed,and some typical routing protocols are summarized and concluded.Secondly,a Topology Aware Routing Protocol Based on Fuzzy Logic and Q-learning(TARP-FQ)is proposed for the scenario with Road Side Units(RSU).Topological changes and network load state are handled by fuzzy logic in TARP-FQ to dynamically adjust the beacon interval.The RSU deployed at intersections can assist in forwarding packets.When the source node or RSU carries packets,the node selects the neighboring RSU with the highest score as the temporary destination node.Each node in the network maintains the Q-value for neighboring RSU and selects the node with the highest Q-value as the next-hop node to forward packets.Moreover,the Q-learning algorithm’s parameters are adjusted according to the link quality and link quality changes in order to update the Q-value and ensure the reliability of the route.Thirdly,a Geographic Routing Based on Vehicle Location Prediction and Node Density(GR-VLPND)is proposed for the scenario without RSU.Based on the GPSR(Greedy Perimeter Stateless Routing),GR-VLPND introduces a location prediction mechanism to update neighboring nodes’ location information and introduces the hybrid beaconing strategy to select the time to send HELLO packets.In addition,the protocol also adjusts the routing strategies according to the node density of the scenario to ensure reliable and efficient packet transmission.Finally,we develop a 3D-VANET simulation platform based on OPNET and Vanet Mobi Sim to evaluate the performance of the two proposed protocols.The simulation results verify the effectiveness of our proposals.