Research On Hierarchical Routing Algorithm Based On Vehicle Distribution In City Vanets

Nowadays,vehicular ad-hoc network(VANET),an important part of Intelligent Transportation System,has gained a great research value and enjoyed a broad prospect.Further,network performance plays a crucial role in the development of VANET among all the research aspects.A sound and reliable routing protocol will facilitate the data transmission and dissemination,thus improving the network performance.However,achieving such a routing protocol will cost great effort due to the fast-changing topology,for nodes of VANET are vehicles and move at a high speed.Aiming at solving the problem,this thesis probes deeply into the VANET routing algorithms of different vehicle spacial distributions.First,when vehicles are densely and evenly distributed on roads in urban area,a location-based hierarchical routing algorithm is proposed,in which road-side units act as gateway and some vehicles are selected as Accessing Help Points(AHPs)according to the density and exact location.The hierarchical routing protocol is proposed with the help of AHPs.In this way,the transmission delay from source vehicle to road-side units would be reduced.Second,when vehicles are sparsely and unevenly distributed on roads in urban area,a headway-based hierarchical routing algorithm is put forward.The vehicles cluster based on the headway in random distribution.The network performance is improved thanks to communication between cluster head and cluster tail.In addition,an improved GPSR routing protocol and hierarchical routing algorithm are both proposed.Moreover,we use simulator Vanet Mobi Sim and NS-2 to conduct a simulation test.It turns out that the location-based hierarchical routing algorithm can effectively reduce delay and packet loss when vehicles are densely and evenly distributed.The headway-based hierarchical routing algorithm suits well on vehicles location change and could enhance the effectiveness of data dissemination.Finally,some problems open to discuss are listed for further research and study.