Routing Protocol Design And Optimization For Vehicular Ad Hoc Networks

VANET (Vehicular Ad hoc Network) is a special kind of MANET (Mobile Ad hoc Network). As one of the most crucial parts of Intelligent Transport Systems (ITS), VANET has been applicated widely in traffic information warning, driving abidance, entertainments and so on.The fundamental component for the success of VANET applications is routing. The unstability of wireless channel and the highly dynamic topology makes the design of routing protocols challenging since it must efficiently handle rapid topology changes and a fragmented network.Aiming at improving the data delivery rate and end-to-end delay of routing in VANETs, a position-and-connectivity-based routing protocol is proposed in this thesis. Besides, a stability-oriented clustering algorithm is proposed; it can improve the broadcast performance in VANETs.Firstly, according to dynamic topology and frequent network disconnections, a probability model of route connectivity is built and a modified Dijkstra algorithm is proposed to calculate the route with the optimal connectivity. Based on the model, a position-and-connectivity-based routing protocol (PCR) is proposed. PCR selects routes with the optimal connectivity so as to improve the packet delivery rate. Moreover, in order to reduce the average delay, vehicles predict on which road segment a neighbouring junction node will forward data to so that the average hop count is decreased. Simulation results under realistic scenarios show that PCR outperforms traditional protocols in packet delivery rate and end-to-end delay.Secondly, according to the large scale of VANET, a stability-oriented weight-based clustering algorithm (SWBCA) is proposed. SWBCA firstly takes into account relative mobility of nodes and the difference between the degree of each node and the ideal degree value for cluster-head election. Then the cluster maintenance of the algorithm is improved with Monte Carlo optimization. Through simulation the SWBCA is proved to be more stable than other algorithms and can help improving broadcast performance through simulation.