Design And Implementation Of Intelligent Intersection Traffic Control System Based On Inter-vehicle Communication

Intersection vehicle control has always been the academic community concernand research hot spots in intelligent transportation systems. Because of the poorsystem performance based on the traditional traffic light, experts try to determine theoptimal time slot of the traffic light in real-time flexible way. However, it is arelatively difficult task in real-time control due to the dynamic of traffic control.With the rapid development of vehicle network, vehicles can communicate andcoordinate managing data with others. In order to avoid the weakness of traditionaltraffic light, we put forward an innovative method based on vehicle communicationcontrol in this paper. Compared to traditional traffic light, we control vehicles viaintersection through exchanging of messages. Unlike conversional vehicle controlmethods, we control traffic by “vehicle” rather than setting the slots of traffic light,which overcome the drawbacks of inefficiency in traditional traffic light.In this work, we redefine intersection vehicular traffic control problems andmodel a new variant of the classic mutual exclusion algorithm. Then, we proposecentralized control algorithm and distributed control algorithm and have proved thealgorithms in theory. Moreover, we ensure the feasibility and correctness of thealgorithm through analyzing some phenomena, such as deadlock, starvation.The author conducts extensive simulations to analyze and evaluate theperformance of our proposed approach based on Ubuntu10.10system and NetworkSimulator-3version. The results show that, our approach is more efficient thantraditional traffic light, such as average waiting time, outputs and so on. We alsodeploy intelligent cars on TinyOS operating system to simulate real-time traffic basedon the centralized control algorithm. The simulating results show that our algorithmscan orderly, coordinately, safely and efficiently control vehicles via the intersection.