Traffic Optimization And Driving Safety Improvement Research Based On Vehicular Ad-hoc Network

Transportation is a vitally important factor for the global economy and the mobility of people.It can bring enormous losses to the economy that affects our social lifestyle.According to WHO records,1.25 million people die,and tens of millions are injured each year because of road traffic crashes,which will make it to become the seventh leading cause of death.In 2012,only in the USA,$121 billion was lost as a result of traffic congestion,including unnecessary fuel consumption and travel delay and expected to reach $199 billion in 2020.To overcome these problems,the transportation system needs to be smart and intelligent enough to assist the drivers and give a warning for the possible accidents as well.Intelligent Transportation System can bring not only convenient,time-saving economic flow and safety transportation for passengers also can reduce time and fuel consumption.The main problems of transportation are traffic safety and congestion mitigation;researchers and car manufacturers are much more interested in researching collision and congestion reduction.Currently,many techniques and technologies are being developed.These systems are partially effective in solving current problems.Nevertheless,most of the existing traffic control system still suffers from several drawbacks and challenges such as high cost,lack of considering drivers’ behaviors,and several factors that can lead to accidents.As the main problem is the high cost to build the communication infrastructure just like servers and the costly roadside units.For the developing countries or big cites with severe traffic congestion,that will cost a massive amount of money to build the communication infrastructure and traffic control system.Another drawback of the existing system is that most of the modern systems neglect driver behaviors and choices.To motivate the drivers to discipline and follow the system rules,the traffic congestion optimization system should consider the factors that can encourage the drivers to choose the optimized paths that can mitigate the traffic congestion.This thesis focuses on traffic safety and traffic congestion mitigation by three different approaches.Firstly,to solve the main challenge of cost-effectiveness,a Distributed Infrastructure-Free Traffic Optimization System(DIFTOS)based on vehicular ad hoc networks for urban environments has been proposed to explain the cost effectiveness of the efficient traffic optimization system for a big city like Beijing.The system does not rely on any centralized server and does not need RSU or any kind of infrastructure.The city map is divided into a hierarchy of servers and the vehicles that are located in the busy road intersections play the role of servers;As the infrastructures and roadside units are the main costly features to make one traffic optimization system,DIFTOS introduces new infrastructure-less hierarchical architecture to cut down the cost.Secondly,the software-based accident prediction system(APS)has been proposed that can consider all the factors which can lead to accidents where most of the APS systems neglected.Such as velocity,weather condition,risky location,nearby vehicles density,and driver fatigue.A Hidden Markov Model(HMM)is used to model the correlation between these observations and the latent variable.Simulation results showed that the proposed system has a better performance in terms of sensitivity and precision compared to the state-of-the-art single factor schemes.Last but not least,a dynamic traffic pricing system by a reward-punishment system to control the congestion has been proposed in which the system sets some rules and policies and dynamic pricing to adapt to peak-hour traffic congestions.The proposed system is implemented using vehicular ad hoc networks,which eliminate the need for installing a costly electronic toll collection system.The system uses newly introduced virtual digital money named T-coin(Traffic Coin),which is used to reward or punish the drivers according to their behaviors.We have proved the effectiveness of the proposed framework through extensive traffic simulations on a real maps and open traffic dataset.The experimental results show that the proposed framework outperforms similar state-of-the-art schemes.