Research On Intersection-Based Routing Protocols In Vehicular Ad Hoc Networks

An Ad Hoc Network consisting of vehicles has emerged and developed as an interesting but challenging domain where lots of new applications may find their place. Such networks are called Vehicular Ad Hoc Networks (VANETs). VANETs are expected to provide applications such as safety applications and user applications, which promotes the development of Intelligent Transportation System (ITS). The research stems from the Natural Science Foundation of China (NSFC) named "Research on the key technology of Vehicular Ad-hoc Networks oriented to the actual scenario and user experience". This research is aimed to proposing an efficient and reliable scheme that is applicable to the urban scenario, and further to promote the standardization of VANETs technology. In VANETs, the end-to-end data transmission is supported by the routing technology, as the various applications in VANETs require more on the networking technology, a reliable and efficient routing is more and more important, so the research of this paper has its practical usage value and significance.What is very important in VANETs is the multi-hop data transmission, which faces new challenges due to the characteristics of VANETs topology. In VANETs, the network topology changes very fast, and traditional topology-based routing protocols cannot be well applied to such high-dynamic topology. Position-based routing is proposed to tackle the problem. To further adapt to the urban scenario, people then propose the intersection-based routing, where the data packets are token priority to transmit to the vehicle located at intersections. However, the current intersection-based routing considers distance and vehicle density as parameters. In fact, it is not sufficient to evaluate the network quality only by vehicle density. The high density does not always signify the high connectivity. Consequently, these intersection-based routing protocols face two problems. The first one is that the connectivity problem due to high vehicle mobility, which causes unreliable vehicle to vehicle (V2V) connection. How to properly select a road segment that is not only close to the destination but also with better connectivity is still not solved. The second is that how to evaluate the network connectivity.In this article, we propose a stochastic traffic model, in which the parameters are set according to the realistic traffic rules. From the perspective of physical connectivity, we take this model as a queuing model. Under the given condition, we deduce the CDF of the adjacent vehicle distance. Besides, we discuss the influence of the adjacent vehicle distance and the communication range on the connectivity, and further give the numerical results. Based on these results, we design the connectivity-sensed routing (CSR) algorithm. Simulation results demonstrate that CSR outperforms the current routing algorithm, especially on the throughput, delivery rate.In the future research of this paper, the traffic flow model will further be improved to add the bi-directional and the intersection model. Besides, with the development of the next generation mobile networks, the assumption about the intersection infrastructure in this paper is expected to be achieved by the small cell in the future 5G.