Design and Implementation of Safety System Model Based on Vehicle-to-Vehicle Communication

Abstract

Recently, studies on safety applications for driver protection are being conducted. However, previous researchers have only surveyed and analyzed traffic accidents from the perspective of traffic engineering methodologies. In addition, advanced driver assistance systems (ADASs) support passive safety systems by including various sensors, such as ultrasonic or vision sensors. However, an ADAS’s high price is an obstacle to popularization and shows low accuracy in bad weather. If vehicular communication is used, the safety system can apply a neighboring vehicle’s information, such as location and speed. In addition, V2V communications systems with low prices are less influenced by weather or obstacles than systems with sensors only. A safety system model using V2V communication has not yet been proposed, nor has a vehicle test been developed and released. In this dissertation, we present the design of a vehicular safety system using V2V communication. The designed vehicular safety system is implemented in a test vehicle and confirmed on a proving ground. This system acquires vehicular location information through a global positioning system (GPS) and receives a neighboring vehicle’s information through an SAE J2735 basic safety message (BSM). The location interpolation method is proposed because location margins are generated due to different periods between GPS and BSM communication. In addition, a gateway for interconnecting a controller area network (CAN) and wireless access in vehicular environments (WAVE) is designed and the performance of the gateway is evaluated. In order to manage information from neighboring vehicles related to safety services, a node management system is proposed. Mathematical models and safety algorithms for forward collision warning (FCW), intersection movement assistance (IMA), left turn assistance (LTA), overtaking assistance (OA), and blind spot warning (BSW) are proposed and implemented in test vehicles. In order to evaluate the system’s performance, test scenarios for the algorithms are configured and the implemented system is tested at the Korea Automotive Technology Institute (KATECH) proving ground. The overall success rate of is 94% and we analyze the field data, such as a reception rate and location errors. The benefits of a V2V communication-based safety service are presented by comparing with the performance of a forward vehicle collision warning system (FVCWS). We describe the communication requirements through network simulation by increasing the number of nodes.