Research On The Theory And Technology Of Traffic Congestion Modeling In Vehicular Ad Hoc Networks

With the development of the automotive industry and wireless technologies, Vehicular Ad Hoc Networks(VANETs) have become the widely studied and the promising areas in recent years.Differing from traditional wired networks, the vehicles in VANETs are equipped with communication devices to enable information sharing between vehicles(V2V) and with the roadside infrastructures(V2I).Therefore, the vehicles in VANETs will obtain the road traffic information through the message exchange within a certain range,whichmay have some impact on their mobile processes, and thus affecting the traffic congestion generate and propagate in this kind of network.Traffic congestion is a major factor in affecting the daily transport systems, which means that the traffic congestion may extend the traveling time and slow down the vehicles’ speeds. So, the travel experiences are greatly reduced. Meanwhile, if too many vehicles arrived to a certain road or street, then the vehicles cannot drive out the actual road timely, and slow down vehicles’ movements or even cause the stagnant phenomenon of this road. In addition, the congestion in one road may affect the other vehicles’ movements in the surrounding roads, in other words, the congestion phenomenon occurs in these roads. Thus the congestion phenomenon spreads throughout the whole networks, which will seriouslyeffect on the performance of the entire networks. Vehicles in VANETs can acquire the traffic information in its surroundings, then the traffic congestion may be mitigated in some degree by changing its traveling route.In VANETs, the communication radius, the traffic densities, traffic lights, mobility strategies(e.g., the shortest time, the shortest distance, or the shortest cost and so on) will affect the traffic congestion. However, most previous studies and researches focused on the unitary impact factor and lack of the comprehensive comparison on the traffic congestion. Many researchers hold the view that the large communication radius will be helpful for the information exchanging among vehicles, and thus the traffic congestion may be mitigated in the large communication radius. Unfortunately, there lacks adeterministic theory modelbetween the traffic congestion and the communication radius so far.This paper is built on the connected vehicles of VANETs, which means that each vehicle can obtain the other vehicles’ position information in a certain degree. We focus on the theoretical model and simulation of traffic congestion by the following aspects:(1) Considering the traffic congestion influenced by different factors in the temporal and spatial preference. These factors include: traffic lights, mobility strategies, traffic densities and communication radius. This paper abstracts the actual traffic system into the lattice system by some steps and introduces some similar concepts to imitate the vehicles’ behaviors in actual roads. We also take two existed optimized schemes into considerations as the comparison to the connected vehicles.(2) We quantitatively deduced the deterministic modelbetween the traffic congestion and the communication radius in the lattice system. Because the traffic congestion can be reflected by the vehicles’ average velocities, i.e. the faster velocity represents a slighter traffic congestion and vice versa. Therefore, deriving the theoretical model about the average velocity and the communication radius is the first step, and then acquire the quantified relations between the traffic congestion and the communication radius further. We adopted the concept of betweenness centrality and queuing theory into the proposed congestionpropagation model in the intermediate deduction process to ensure the validity and the rationality of our model.Through running the theoretical model and the corresponding movement rules on the simulation software Matlab, the expected results are all confirmed. The experiment results show that the traffic lights, traffic densities, move patterns and the communication radius are all capable to effect on the traffic congestion. However, the communication radius fails to reduce the traffic congestion consistently when reached to the optimal value, which indicates that the traffic congestion can be mitigated mostly in an optimal communication radius. Through the comparison between the existed optimized algorithms and the connected vehicles, we can conclude that the V2 V communications in VANETs will accelerate the vehicles’ average vehicles to achieve lower traffic congestion. Totally, this paper will be beneficial to understand the congestion generation and propagation in VANETs.