在十字路口常會有駕駛搶黃燈或闖紅燈的行為，導致在經過路口時與另一向遵守交通號誌的駕駛發生碰撞的車禍，本論文提出一利用都普勒偏移特性偵測路口碰撞行為的方法，因兩車的移動方向皆相同，若在觀察車等速的條件下，利用都普勒偏移公式可推得，如果都普勒偏移值一直維持固定，則兩車在路口很有可能會發生碰撞。本論文皆是利用車對車間發射無線電波的都普勒特性進行行車安全的預警分析，以達到低延遲快速預警的效果，這也是V2V通訊碰撞預警中最重要的一環。綜合以上的研究項目，在未來5G V2V通訊系統的情境下，可為行車駕駛的生命安全提供迅速，有效的安全資訊，並有效的減少車禍的發生。 In this dissertation it explores and develops a collision avoidance process for Vehicle-to-Vehicle communication in Non-Line-of-Sight (NLOS) communication environment for 5G mobile communication networks. It consists of three chapters in the dissertation; in Chapter 2 the scenario simulation and analysis, it considers in freeway three vehicles moving environments and then uses simulations to observe the variations of vehicles Doppler frequency shifts in every environment; in Chapter 3 a vehicle collision theory is proposed and a vehicle collision pre-warning process is developed to determine is there any vehicle collision occurs and then in Chapter 4 it is through simulations to validate our proposed collision pre-warning algorithm.
It is reported quite often in daily news that serious rear-end car collisions are happened in Taiwan freeway due to drivers inappropriate driving behavior or bad view range due to weather condition. In current prosperously developed IoT (Internet of Things) and 4G/5G wireless communication networks car industry has invested many resources in the development of car collision safety technologies trying to possibly reduce property damages and fatalities. In this dissertation it considers in the freeway traffic environment to have cars installed V2V transceiver to utilize the Doppler frequency drift characteristic and to analyze the frequency drift generated from moving vehicles; and then to observe the dynamic relationships among vehicles and through the variation of frequency drift to detect the situation when many in front vehicles are moving in the same lane to perform the pre-warning analysis for possible vehicles collision and also to analyze is it possible to safely complete the car pass intention when a vehicle is trying, across from another lane, to pass a front vehicle moving in the same lane.
In a cross-section road it sometimes occurs the situation a defense driver is hit at the intersection by a vehicle that its driver does not stop at red signal or rushes to ignore the yellow signal. In this dissertation we propose a method to utilize the Doppler frequency drift characteristic to detect vehicle collision behavior at the intersection. If constant speed vehicles are moving in the same direction and if the Doppler frequency drift is maintained at a fixed value then by utilizing the developed Doppler frequency drift formula it could detect the possibility that two vehicles may incur collision at the intersection
In this dissertation it utilizes the Doppler frequency shift characteristic of wireless electromagnetic waves emitted between vehicles to perform the pre-warning analysis for the safety of moving vehicles it attains a low delay and fast pre-warning effect that is also the most important subject in considering the collision pre-warning issue in vehicle–to-vehicle communication. From the subjects we analyzed and developed in this dissertation it concludes that we can provide a vehicle driver a fast and safety information to effectively reduce the occurrence of possible fatalities.
